Why Exposure Needs a Measure
An insurer may cover one vehicle for a year, several buildings for different parts of a year, or a workforce whose payroll changes over a policy period. Simply counting policies or insured objects does not show how much risk was covered: one policy might cover one object for a month, while another covers many objects for a full year. An exposure unit provides a stated basis for measuring the amount of insured risk.
The basis depends on the coverage and the question being asked. Vehicle-years and location-years combine a count of insured items with time. Payroll and insured property value measure a different dimension of exposure. A meaningful exposure figure must therefore include its units; “500” alone is incomplete, while “500 vehicle-years” identifies what was measured.
An exposure measure describes how much risk was subject to coverage, not whether a loss occurred or how large it was. For example, a fleet might have 10 vehicle-years of exposure and three claims during that time. The exposure is 10 vehicle-years; the three claims are a separate observation. Keeping numerator and denominator concepts distinct is essential whenever data are summarized.
Time-Based Exposure
For a time-based measure, each insured item contributes exposure in proportion to the time it was covered, provided the chosen unit treats each item equally. One vehicle covered for a full year contributes one vehicle-year. One vehicle covered for six months contributes one-half of a vehicle-year. If the covered time is measured in months, divide the number of covered months by 12 to express exposure in years.
This addition follows from the meaning of the unit. If an item is covered over two non-overlapping periods, its total covered time is the sum of the period lengths. Across multiple items, add the covered time for each item. Thus, if the same number of items is covered throughout each interval, the amount of exposure is that number multiplied by the interval length, and the total is the sum across intervals.
where \(e\) is exposure in item-years, \(w_j\) is the number of insured items in group \(j\), and \(t_j\) is the fraction of a year each item in that group was covered. This is an accounting identity for the stated exposure basis. It does not assert that the chance of loss is constant over time, or that every item has the same underlying risk.
Worked Example: Fleet Exposure With Partial-Year Coverage
A hypothetical insurer covers eight delivery vans for a full 12-month year, three more vans for the final six months, and two additional vans for the final three months. During the year, three claims are reported. Calculate the fleet’s exposure in vehicle-years and state what the claim count means in relation to that exposure. Assume each listed van is covered only for the stated period, with no overlapping duplicate entries. No probability distribution is assumed.
Method. Convert each coverage period to a fraction of a year, multiply by the number of vans, then add the groups. Keep the observed claim count separate from the exposure calculation.
The first group contributes \(8 \times 12/12=8\) vehicle-years. The second contributes \(3 \times 6/12=1.5\) vehicle-years. The third contributes \(2 \times 3/12=0.5\) vehicle-years.
The three reported claims do not change the exposure: it remains 10 vehicle-years. If the observed claim count is divided by exposure for descriptive purposes, the result is \(3/10=0.3\) claims per vehicle-year. That ratio describes these hypothetical observations; it is not itself an exposure unit and does not establish a probability model.
Check. The exposure groups represent eight full-year van equivalents, 1.5 full-year equivalents, and 0.5 full-year equivalents. Their sum is 10 vehicle-years. The common mistake this calculation avoids is adding the number of vehicles (13) without accounting for the shorter coverage periods.
Choosing and Naming the Basis
A count-and-time basis is useful when the exposure measure treats each insured item as one unit. It does not capture differences in size or value between items. For other purposes, exposure may be measured by an amount such as payroll or insured property value. In those cases, state the amount and the period clearly. A total payroll figure for a year is not interchangeable with a count of employees or employee-years; likewise, a dollar amount of property is not interchangeable with a count of locations.
The coverage context determines which basis is meaningful. Some common forms are shown below. These are examples of measurement bases, not claims that every policy of a given type must use that basis.
| Exposure basis | What is counted or measured | Example unit |
|---|---|---|
| Vehicle-time | Vehicles and the time each is covered | Vehicle-year |
| Location-time | Covered locations and the time each is covered | Location-year |
| Payroll | Payroll during a specified coverage period | $1,000 of payroll during the period |
| Insured property value | Covered property value during a specified period | $1,000 of insured value for one year |
Worked Example: Payroll Is Not an Employee Count
A hypothetical business has two work groups covered for the same full policy year. The first group has $480,000 of payroll during that year, and the second has $320,000. The stipulated exposure basis is one unit for each $1,000 of payroll during the policy year. Calculate total exposure units. Also explain why the number of employees cannot be substituted for the payroll amounts. Assume the stated payroll figures relate to this same one-year coverage period; no probability distribution is assumed.
Method. Add payroll for the common period, then convert dollars to the stated $1,000 unit.
Combined payroll is \(480{,}000+320{,}000=800{,}000\) dollars. Dividing by $1,000 per unit gives \(800{,}000/1{,}000=800\) exposure units.
The result is 800 units of $1,000 of payroll during the stated year. It is not 800 employees, 800 employee-years, or 800 dollars. Employee counts cannot replace the payroll basis because two groups with the same number of employees could have different total payroll.
Check. Multiplying 800 units by $1,000 per unit recovers the combined payroll of $800,000. The common mistake this avoids is reporting the dollar total as though it were already a count of exposure units.
Value-Based Exposure and Partial Periods
When the basis combines property value and time, each property contributes its insured value multiplied by the fraction of the year it was covered. The resulting units can be described as dollar-years, or as thousands of dollars of insured value-years if values are recorded in thousands. By contrast, a location-year gives each covered location equal weight, regardless of its insured value. These two measures answer different questions and should not be added together.
Worked Example: Location-Years and Insured-Value-Years
A hypothetical schedule covers Property A, valued at $240,000, for a full year; Property B, valued at $360,000, for a full year; and Property C, valued at $120,000, for nine months. Assume all three are covered continuously for exactly the stated periods, their insured values stay constant while covered, and there are no overlapping duplicate entries. Calculate both total location-years and total insured-value exposure in thousands of dollars for each year of coverage. No probability distribution is assumed.
Method. For location-years, count each location’s covered time. For value-years, multiply each property’s value, expressed in thousands of dollars, by its covered fraction of a year. Do not confuse the two bases.
Property A contributes \(12/12=1\) location-year, and Property B also contributes \(12/12=1\) location-year. Property C contributes \(9/12=0.75\) location-year. Therefore:
For the value-based measure, the insured values in thousands of dollars are 240, 360, and 120. Property A contributes \(240 \times 1=240\) thousand-dollar-years. Property B contributes \(360 \times 1=360\) thousand-dollar-years. Property C contributes \(120 \times 9/12=90\) thousand-dollar-years. Thus:
Check. The location total is \(1+1+0.75=2.75\), while the value total is \(240+360+90=690\) thousand-dollar-years. Each full-year property contributes exactly one location-year, not more than one; the nine-month property contributes three-quarters of one. The common mistake this calculation avoids is using a property’s value to change the number of location-years or adding location-years to value-years.
Common Exposure Errors
Exposure calculations often involve straightforward arithmetic, but the interpretation can still go wrong. Before reporting a total, identify the basis, identify the covered period, and check that every component is expressed in compatible units.
- Counting claims as exposure. Exposure is the amount of risk measured on the chosen basis. Claims are observations that may be compared with exposure, but they are not part of the exposure total.
- Ignoring partial coverage. A vehicle covered for three months contributes \(3/12\) of a vehicle-year, not one vehicle-year, if the chosen basis is vehicle-years.
- Using the wrong time denominator. A period of six months is \(6/12=0.5\) years, not 6 years. Use a time conversion consistent with the unit being reported.
- Combining unlike measures. Vehicle-years, payroll units, location-years, and insured-value-years are different units. They cannot be summed into a meaningful total without a defined conversion.
- Assuming an exposure unit guarantees equal risk. Equal vehicle-years do not necessarily mean equal hazard. Exposure is a measurement basis; it does not, by itself, describe the chance or severity of loss.
Decide whether the stated measure is a count-and-time unit, payroll, insured value, or another specified basis.
Convert time fractions or amounts so all terms use the same unit and coverage period.
Add compatible contributions, then report the result with its unit. Keep claim counts and loss amounts separate.
The distinction matters because an exposure figure is only interpretable when its basis is explicit. “2.75” by itself does not tell a reader whether the total is location-years or something else; “2.75 location-years” does. A carefully named denominator also makes later comparisons possible without mistaking the amount of insured risk for the claims observed against it. This foundational measurement skill is aligned with the SOA Exam FAM (Fundamentals of Actuarial Mathematics) syllabus.
Check Your Understanding
For each problem, identify the exposure basis and show how partial periods or measurement units affect the total.
- A hypothetical insurer covers four motorcycles for a full year and two motorcycles for six months. What is the total exposure in motorcycle-years? (Answer: 5 motorcycle-years)
- A hypothetical shop has $210,000 of payroll during a one-year coverage period. If the specified unit is $1,000 of payroll during that period, how many exposure units does it have? (Answer: 210 units)
- A hypothetical schedule covers five storage locations for a full year and one additional location for four months. Calculate total location-years. (Answer: \(5+4/12=5\frac{1}{3}\) location-years)
- A hypothetical fleet has 12 vehicles covered for a full year and four vehicles covered for the first three months only. If five claims are observed, state the vehicle-year exposure and distinguish it from the claim count. (Answer: \(12+4(3/12)=13\) vehicle-years; the five claims are a separate observed count.)
- A hypothetical property schedule covers a $150,000 building for one full year and a $250,000 building for six months. Using thousands of dollars of insured value-years, calculate total exposure. (Answer: \(150(1)+250(6/12)=275\) thousand-dollar-years)