Worked Expected Loss Cost Calculations

Why Exposure Units Matter

An expected loss cost is meaningful only in relation to a defined exposure unit. A cost stated per policy-year, per $1,000 of payroll, or per 1,000 square-foot-years has a different denominator in each case. To estimate expected ultimate losses, first express the exposure on the same basis as the loss cost; then multiply. A mismatch in units can produce an answer that is wrong by a factor of 1,000 or more even when the arithmetic is otherwise correct.

This tutorial focuses on calculating the expected ultimate loss from an exposure count and an expected loss cost. The exposure count may be a direct count of unit-years, a count of payroll-based units, or a converted measure such as insured area over time. In every case, the loss cost must represent the expected ultimate loss for exactly the exposure unit being counted.

Definition: An exposure unit is the unit to which an expected loss cost applies, such as one policy-year or $1,000 of payroll. An expected loss cost is the expected ultimate loss associated with one specified exposure unit. If $N$ is the number of matching exposure units and $e$ is the expected ultimate loss cost per unit, the exposure-based expected ultimate loss estimate is $N e$.

The word ultimate matters: the selected loss cost is an estimate of all losses ultimately attributable to the exposure, not merely the amount paid so far. The result is an expected total for the covered period and loss scope specified in the problem. It is not, by itself, an estimate of unpaid loss.

The Exposure-Based Calculation

Let $U$ denote the aggregate ultimate loss for a group of exposures. Suppose there are $N$ comparable exposure units, and each unit has expected ultimate loss $e$. The expected aggregate loss is

$$ E[U] = N e. $$

The multiplication is valid when $e$ is expressed per unit of the same exposure counted in $N$, and when the cost covers the same period, coverage, and loss scope. For example, a loss cost per policy-year cannot be multiplied directly by a count of policies that were insured for different fractions of a year. Convert those policies to policy-years first.

Why the Formula Holds

The formula follows from adding the losses for the individual exposure units. Let $X_j$ be the ultimate loss associated with exposure unit $j$, for $j=1,\ldots,N$. The aggregate loss is the sum of those unit losses. If each unit has the same expected loss $e$, linearity of expectation gives the result.

U = \sum_{j=1}^{N} X_j \qquad\Longrightarrow\qquad E[U] = E\left[\sum_{j=1}^{N} X_j\right] = \sum_{j=1}^{N} E[X_j] = \sum_{j=1}^{N} e = N e.

This derivation does not require the losses from different exposure units to be independent. It requires that the expected loss for each counted unit is the stated $e$, and that the aggregate is the sum of those losses. If exposures have different expected costs, their expected losses can be calculated separately and added; applying one common cost to unlike units would need additional justification.

Identify the exposure unit.
Read the denominator in the loss-cost statement: one policy-year, $1,000 of payroll, or another specified unit.

Express the exposure in those units.
Convert partial terms, payroll amounts, area, or other measures as needed. Keep track of the conversion factor.

Multiply by the loss cost.
Multiply the number of matching units by the expected ultimate loss cost per unit.

Check the result.
Confirm that the exposure units cancel, the answer has units of dollars, and its scale is plausible for the stated data.

Worked Calculations

Worked Example: Converting Partial Terms to Policy-Years

A hypothetical equipment coverage has an expected ultimate loss cost of $214.60 per equipment policy-year. During the period, 1,200 policies are in force for a full year and 500 additional policies are in force for six months each. Estimate the expected ultimate loss.

Method. Convert all insured terms to policy-years, then multiply by the loss cost per policy-year. A full-year policy contributes 1 policy-year; a six-month policy contributes $6/12=0.5$ policy-year.

The full-year policies contribute $1,200 \times 1 = 1,200$ policy-years. The partial-year policies contribute $500 \times 0.5 = 250$ policy-years. Therefore, total exposure is $1,200+250=1,450$ policy-years.

Multiply by the expected loss cost: $1,450 \times \$214.60 = \$311,170.00$. The estimate is $311,170.00.

Check and common mistake avoided. The units are policy-years multiplied by dollars per policy-year, leaving dollars. Counting all 1,700 policies as full-year exposures would give $1,700 \times \$214.60=\$364,820.00$, overstating the estimate by treating each six-month policy as a full year. The correct calculation reflects the actual insured duration.

Worked Example: Loss Cost per $1,000 of Payroll

A hypothetical trade policy has an expected ultimate loss cost of $73.00 per $1,000 of payroll. The covered payroll for the period is $48,600,000. Estimate expected ultimate losses.

Method. Convert total payroll into $1,000 units. The exposure count is $\$48,600,000 \div \$1,000 = 48,600$ payroll units.

Each unit carries an expected loss cost of $73.00, so the estimate is $48,600 \times \$73.00$. Breaking down the multiplication, $48,600 \times 70 = 3,402,000$ and $48,600 \times 3 = 145,800$. Their sum is $\$3,547,800.00$.

The expected ultimate loss estimate is $3,547,800.00. As a check, the $1,000 payroll units cancel against the loss cost’s “per $1,000 of payroll” denominator, leaving dollars.

Common mistake avoided. Multiplying $48,600,000 directly by 73 treats the rate as $73 per dollar of payroll, not $73 per $1,000. That incorrect calculation gives $3,547,800,000, which is 1,000 times the correct result. Dividing payroll by 1,000 before applying the rate—or equivalently multiplying payroll by $73/1,000$—preserves the stated scale.

Worked Example: Area Exposure With a Partial-Year Term

A hypothetical property program has an expected ultimate loss cost of $18.40 per 1,000 square-foot-years of insured area. It insures 950 buildings, each averaging 2,400 square feet, for ten months during the period. Estimate expected ultimate losses.

Method. First calculate square-foot-years, accounting for the ten-month term. Total insured area is $950 \times 2,400 = 2,280,000$ square feet. Ten months is $10/12$ of a year, so the exposure is $2,280,000 \times 10/12 = 1,900,000$ square-foot-years.

The loss cost is stated per 1,000 square-foot-years. Convert the exposure to those units: $1,900,000 \div 1,000 = 1,900$ units. Then calculate $1,900 \times \$18.40$. The parts are $1,900 \times 18 = 34,200$ and $1,900 \times 0.40 = 760$, totaling $\$34,960.00$.

The expected ultimate loss estimate is $34,960.00. The dimensions check: square feet multiplied by years, divided by 1,000 square-foot-years per unit, gives the number of units; multiplying by dollars per unit gives dollars.

Common mistake avoided. Using 2,280,000 square feet as though it were 2,280,000 units ignores both the ten-month duration and the 1,000-square-foot scale. The correct exposure is 1,900 loss-cost units, not the raw area in square feet.

Reading the Units and Interpreting the Result

Writing the units alongside the numbers is a practical safeguard. For a loss cost per $1,000 of payroll, the calculation can be displayed as

\left(\frac{\$48{,}600{,}000\ \text{payroll}}{\$1{,}000\ \text{payroll per unit}}\right) \left(\frac{\$73\ \text{loss}}{\text{unit}}\right) = \$3{,}547{,}800\ \text{expected loss}.

This check is especially useful when a problem supplies several scales at once, such as thousands of dollars, millions of dollars, partial years, or rates per a stated number of exposure units. Convert one scale at a time and retain enough precision through the calculation. Round the final money amount to cents when the problem calls for dollars and cents; avoid rounding a converted exposure prematurely if it would affect the final answer.

Expected loss is an average over the uncertainty in future losses, not a promise that actual losses will equal the estimate. An actual outcome may be larger or smaller. The formula estimates an expected total only under the stated exposure and expected-loss-cost basis; it does not by itself establish that the cost is appropriate for a different coverage, time period, or exposure definition.

Common Exam Traps

Using policies instead of policy-years. A policy in force for half a year contributes half a policy-year when the rate is per policy-year.
Ignoring a rate’s scale. “Per $1,000” means convert the exposure to thousands of dollars, not multiply the full dollar amount by the quoted cost.
Mixing area and area-time. If the unit is square-foot-year, both the area and the time component must be represented.
Confusing ultimate loss with paid loss. The selected expected loss cost estimates ultimate loss. Paid-to-date amounts are not substituted for that cost in the exposure-based formula.
Rounding before conversion is complete. Carry the exact exposure-unit count through multiplication, then round the final monetary answer as instructed.

Key takeaway: Expected ultimate loss equals the number of exposure units multiplied by expected ultimate loss cost per matching unit. The most important check is dimensional: exposure units must match the loss-cost denominator, so that the product is a monetary amount.

Check Your Understanding

For each problem, convert the exposure to the stated loss-cost basis before multiplying. Answers are shown so you can check both the unit conversion and the arithmetic.

A hypothetical service policy has an expected ultimate loss cost of $240 per policy-year. There are 920 full-year policies. What is the expected ultimate loss? (Answer: $220,800.00)
A hypothetical coverage has an expected loss cost of $6.80 per $1,000 of payroll. Covered payroll is $12,600,000. Estimate expected ultimate losses. (Answer: $85,680.00)
A hypothetical warehouse program has an expected loss cost of $465 per 10,000 square-foot-years. It insures 1,750,000 square feet for one year. Estimate expected ultimate losses. (Answer: $81,375.00)
A hypothetical vehicle program has an expected loss cost of $1,180 per 100 vehicle-years. Its exposure is 3,250 vehicle-years. Calculate the expected ultimate loss. (Answer: $38,350.00)
A hypothetical insurer has two exposure groups with the same loss scope: 740 unit-years at $315 per unit-year and 260 unit-years at $480 per unit-year. Calculate each group’s expected ultimate loss and then the combined estimate. (Answer: $233,100.00 and $124,800.00; combined expected ultimate loss $357,900.00)

What You'll Learn