Updating Expected Losses for Exposure Changes

Why Expected Losses Must Match the Exposure Record

An expected loss benchmark is meaningful only when its exposure amounts, classifications, coverage, and rates describe the same risk basis. If an audit changes payroll, or a review changes which class code applies, the expected losses used in an experience plan may need to be recalculated. A changed benchmark does not, by itself, mean that the underlying expected-loss rates have changed.

The practical task is to identify what changed. A correction to payroll changes the measured exposure amount. A correction to a class code can change which rate applies, even when total payroll is unchanged. A change in coverage can require a rate for a different coverage basis. These effects should be separated so the recalculation can be explained and checked.

Definition: For a specified classification and coverage basis, an expected-loss rate is the expected loss dollars per stated exposure unit. If $E_j$ is the number of exposure units in group $j$, and $r_j$ is the matching expected-loss rate per unit, then the expected loss for that group is $E_jr_j$. Total expected losses are the sum across the relevant groups.

The exposure unit might be a fixed amount of payroll, one vehicle-year, or another defined measure. Convert recorded exposure to the rate's unit before multiplying. For example, if payroll is recorded in dollars and the rate is dollars per $100 of payroll, the exposure units are payroll divided by 100. Do not multiply raw payroll dollars by a rate stated per $100.

The rate must also match the group to which it is applied. In this tutorial, regard the applicable rates as already selected for the relevant experience period and stated classification and coverage basis. The recalculation updates the exposure record and rate assignment; it does not independently develop new rates.

A Reconciliation Framework

Let $j$ index groups defined by the applicable classification, coverage, and any other rate-basis distinctions. Let $E_j^{\text{old}}$ and $E_j^{\text{new}}$ be the old and revised exposure units, and let $r_j$ be the rate per exposure unit for that group. If the same rate basis applies in both calculations, then:

L^{\text{old}}=\sum_j E_j^{\text{old}}r_j, \qquad L^{\text{new}}=\sum_j E_j^{\text{new}}r_j, \qquad \Delta L=L^{\text{new}}-L^{\text{old}}.

If the rate basis itself changes, define the old and new groups and rates separately. For example, coverage may have changed so that some exposure now belongs with a different coverage-specific rate. The new benchmark is then calculated using the revised exposure assigned to each applicable new group. Do not retain the old rate merely because the overall exposure measure, such as payroll, is unchanged.

Confirm the exposure measure and unit.
Identify what is counted and the rate denominator, such as payroll dollars converted to units of $100.

Reconcile the exposure records.
Use the corrected or audited amounts and determine whether the change affects total volume, the allocation among groups, or both.

Assign each amount to its rate basis.
Match classification and coverage to the applicable expected-loss rate. Keep distinct groups separate when their rates differ.

Calculate and check.
Multiply units by their matching rates, sum the group results, and reconcile the difference from the previous benchmark.

Why the Multiplication and Summation Work

Suppose an expected-loss rate $r_j$ is defined as expected loss dollars per exposure unit for group $j$. For $E_j$ comparable units, multiplying the per-unit expectation by the number of units gives the aggregate expectation. In particular, if each unit has the same expected loss $r_j$, linearity of expectation gives:

E\left[\sum_{k=1}^{E_j} X_{jk}\right] =\sum_{k=1}^{E_j}E[X_{jk}] =\sum_{k=1}^{E_j}r_j =E_jr_j.

This argument does not require the losses for different units to be independent; linearity of expectation is sufficient. It does require the rate to represent the expected loss per unit for the units being counted. If units differ materially in classification or coverage, group them and use the appropriate rate for each group rather than treating all units as interchangeable. Summing the group expectations then gives the total expected loss.

Worked Example: Updating Payroll After an Audit

A hypothetical employer initially reports $1,200,000 of payroll in one classification. The expected-loss rate is $2.40 per $100 of payroll. An audit establishes that the correct payroll for the experience period is $1,350,000, with no change in classification or coverage. Recalculate the expected losses and reconcile the change. State the method and a common mistake to avoid.

Method. Convert each payroll amount to units of $100, then apply the same classification- and coverage-specific rate. Because only payroll volume changed, the rate assignment stays fixed.

The original exposure is:

E^{\text{old}}=\frac{\$1{,}200{,}000}{\$100}=12{,}000\text{ units}.

The original expected losses are:

L^{\text{old}}=12{,}000\times \$2.40=\$28{,}800.

The audited exposure is:

E^{\text{new}}=\frac{\$1{,}350{,}000}{\$100}=13{,}500\text{ units}.

The revised expected losses are:

L^{\text{new}}=13{,}500\times \$2.40=\$32{,}400.

The change is $ \$32{,}400-\$28{,}800=\$3{,}600$. As a check, the audit added $150,000 of payroll, or 1,500 exposure units, and $1{,}500\times \$2.40=\$3{,}600$. The two calculations agree.

Common mistake. Multiplying $1,350,000 by $2.40 would produce an incorrect result because the rate is per $100 of payroll, not per dollar of payroll. The corrected calculation uses 13,500 rate units.

Classification Changes Versus Total Exposure Changes

A revised class-code assignment can change the expected loss benchmark even when total exposure is exactly the same. In that case, the total amount of exposure has not increased or decreased; rather, exposure has moved between rate groups. The change in expected losses reflects the different rates attached to those groups.

Worked Example: Reallocating Payroll Between Classifications

A hypothetical business reports $400,000 of payroll in Class A and $200,000 in Class B. The rates are $1.80 and $4.50 per $100 of payroll, respectively. A classification review determines that the correct allocation is $250,000 to Class A and $350,000 to Class B. Total payroll remains $600,000. Recalculate expected losses and identify whether the change is due to exposure volume or rate assignment.

Method. Calculate each class separately in units of $100. Do not apply a single rate to total payroll, because the class-specific rates differ.

Before the review, Class A has $ \$400{,}000/\$100=4{,}000$ units, giving $4{,}000\times \$1.80=\$7{,}200$. Class B has $ \$200{,}000/\$100=2{,}000$ units, giving $2{,}000\times \$4.50=\$9{,}000$. Thus:

L^{\text{old}}=\$7{,}200+\$9{,}000=\$16{,}200.

After the review, Class A has $ \$250{,}000/\$100=2{,}500$ units, giving $2{,}500\times \$1.80=\$4{,}500$. Class B has $ \$350{,}000/\$100=3{,}500$ units, giving $3{,}500\times \$4.50=\$15{,}750$. Thus:

L^{\text{new}}=\$4{,}500+\$15{,}750=\$20{,}250.

The revised benchmark is higher by $\$20{,}250-\$16{,}200=\$4{,}050$. Total payroll remains $600,000 both before and after. The change results from moving 1,500 units from the lower-rate class to the higher-rate class: $1{,}500\times(\$4.50-\$1.80)=\$4{,}050$.

Check and common mistake. The revised class amounts sum to $250,000 + $350,000 = $600,000, confirming total payroll is unchanged. Treating this as a mere increase in total payroll would misstate the reason for the increase; applying either class rate to all payroll would also be incorrect.

Coverage Changes and Rate-Basis Matching

Coverage affects which expected-loss rate is appropriate when the rates are defined for different coverage provisions, such as different limits or forms. A coverage change may leave payroll or another exposure measure unchanged while changing the rate basis. In that situation, the calculation must assign each part of the exposure to the coverage it actually had, using the matching rate.

This requires reliable information about the exposure associated with each coverage basis. If the record does not identify which exposure belongs to which coverage, the calculation cannot be made correctly from total exposure alone. The assumed allocation must be clarified rather than silently applying one rate to all exposure.

Worked Example: Reassigning Exposure After a Coverage Change

A hypothetical insured has $900,000 of payroll. The original expected-loss calculation treated all payroll as having a $1 million limit, for which the rate is $3.20 per $100 of payroll. A coverage review shows that $500,000 had the $1 million limit and $400,000 had a $2 million limit. The rate for the $2 million limit is $3.75 per $100. Recalculate expected losses and separate the effect of the coverage basis from the total payroll amount.

Method. Keep the two coverage groups distinct and apply each coverage-specific rate. The total payroll remains $900,000; only its allocation across rate bases is corrected.

Under the original assignment, payroll units were $ \$900{,}000/\$100=9{,}000$. The original expected losses were:

L^{\text{old}}=9{,}000\times \$3.20=\$28{,}800.

For the revised assignment, the $1 million-limit group has $ \$500{,}000/\$100=5{,}000$ units, so its expected losses are $5{,}000\times \$3.20=\$16{,}000$. The $2 million-limit group has $ \$400{,}000/\$100=4{,}000$ units, so its expected losses are $4{,}000\times \$3.75=\$15{,}000$. Therefore:

L^{\text{new}}=\$16{,}000+\$15{,}000=\$31{,}000.

The increase is $\$31{,}000-\$28{,}800=\$2{,}200$. The check is that the revised payroll groups sum to $500,000 + $400,000 = $900,000. The expected-loss change can also be checked as $4{,}000\times(\$3.75-\$3.20)=\$2{,}200$, since only the 4,000 units assigned to the higher-limit rate differ from the original basis.

Common mistake. Applying the $3.75 rate to all 9,000 units would treat the entire payroll as having the $2 million limit and produce $33,750. That contradicts the coverage record. The calculation must follow the exposure allocation, not just the overall payroll total.

Checks That Prevent Basis Errors

Before finalizing an updated expected-loss calculation, check both arithmetic and meaning. A numerically correct product can still be wrong if it uses the wrong class code, coverage rate, or exposure unit. These checks are particularly useful when an audit changes multiple fields in the same record.

Check units. Confirm that exposure and rate use the same denominator. Convert payroll dollars to payroll units when the rate is per $100.
Check totals. Reconcile exposure across all revised groups to the audited total. Reclassification or coverage allocation should not create or discard exposure.
Check rate assignments. Confirm that each group's rate matches its classification and coverage basis.
Check the source of change. State whether expected losses changed because total exposure changed, because exposure moved between rate groups, or because both occurred.
Check for double counting. When exposure has been split into groups, sum the group calculations once; do not also include the unsplit total as another group.

A useful reconciliation is to report the old benchmark, the revised benchmark, and the difference, together with the exposure and rate basis used for each. This makes it possible to distinguish a genuine change in measured exposure from a corrected classification or coverage assignment. The distinction matters because expected losses are a benchmark constructed from exposure and assumptions; they are not the actual losses and do not change merely because actual claims changed.

Key takeaway: Recalculate expected losses by converting exposure to the rate's unit, assigning each amount to its correct classification and coverage basis, and summing exposure times the matching rate. A changed total benchmark can reflect more exposure, a different mix of rate groups, or both.

Check Your Understanding

For each problem, use the stated rate basis and check whether the change is in exposure volume, allocation, or both.

A hypothetical account's audited payroll is $780,000, compared with a previously reported $720,000. The unchanged expected-loss rate is $2.75 per $100 of payroll. What are the old and revised expected losses and the change? (Answer: Old $19,800; revised $21,450; increase $1,650.)
A business originally assigned all $1,100,000 of payroll to Class X at $1.60 per $100. A review instead assigns $700,000 to Class X and $400,000 to Class Y, whose rate is $3.10 per $100. Calculate the revised expected losses and the change from the original calculation. (Answer: Revised $23,600; increase $6,000 from the original $17,600.)
A vehicle fleet's exposure is corrected from 2,400 vehicle-years to 2,250 vehicle-years. The rate is $18.50 per vehicle-year. Calculate the original expected losses, revised expected losses, and decrease. (Answer: Original $44,400; revised $41,625; decrease $2,775.)
A hypothetical policy's $800,000 payroll was originally all assigned to a basic coverage rate of $2.20 per $100. The corrected record assigns $300,000 to basic coverage and $500,000 to enhanced coverage, whose rate is $2.90 per $100. Calculate the revised expected losses and the change. (Answer: Revised $21,100; increase $3,500 from the original $17,600.)
Payroll is $640,000 and the applicable expected-loss rate is $4.20 per $100 of payroll. Calculate expected losses, and explain why multiplying $640,000 directly by $4.20 is not valid. (Answer: $26,880; the exposure is 6,400 units of $100, so $6{,}400\times \$4.20=\$26{,}880$.)

What You'll Learn