Midterm Endorsements And Exposure Changes

Why Midterm Changes Require an Exposure Timeline

A policy’s exposure can change before its term ends. A hypothetical endorsement might add a covered vehicle, change a property limit, or remove a covered item. For ratemaking, the key question is not how many transactions appear in the policy records. It is what coverage was in force during each part of the term. A timeline answers that question and prevents the same coverage from being counted twice.

The appropriate calculation depends on the exposure base for the coverage. If exposure is measured in unit-years, use the number of covered units during each interval. If it is measured by insured amount, weight each interval by the amount in force. Do not substitute one basis for the other: first identify the exposure unit, then apply it consistently.

Definition: A midterm exposure interval is a portion of a policy term during which the relevant exposure quantity remains constant. An effective change creates a new interval. The intervals used for a calculation must be successive and non-overlapping, and together must cover exactly the period being measured.

Building Non-Overlapping Intervals

Start with the policy term and list every effective date on which the quantity relevant to the exposure base changes. Split the term at those dates. For instance, if an insured amount changes on day 101, the period before the change and the period beginning on day 101 are separate intervals. A change on a boundary does not create an extra day: treat intervals as adjoining, with the change effective at the stated boundary.

For a term measured in days, let $n_j$ be the number of days in interval $j$, and let $q_j$ be the number of covered units in force during that interval. If one unit-year is 365 unit-days, the time-based exposure is:

E_{\text{units}}=\sum_j q_j\frac{n_j}{365}.

The denominator must match the convention specified for the calculation. If the problem states a different year basis, use that basis instead. For amount-based exposure, let $A_j$ be the insured amount in force during interval $j$, and let $K$ be the amount that defines one exposure unit. Then:

E_{\text{amount}}=\sum_j \frac{A_j}{K}\frac{n_j}{365}.

These are sums over time segments, not over policy transactions. A record that describes the original policy amount for a full term may be superseded in part by an endorsement. When calculating exposure, use the amount actually in force on each day, not the full-term amount from every record.

Choose the exposure base.
Determine whether the calculation calls for covered unit-time, insured-amount-time, or another specified basis.

List effective changes.
Use the effective dates of changes that affect the chosen exposure quantity, rather than counting each record as a separate full-term exposure.

Partition the term.
Write successive intervals that cover the measurement period once, with no gaps or overlaps, and record the active quantity in each.

Weight and add.
Multiply each interval’s duration by its active unit count or normalized insured amount, then divide by the stated time basis.

Check the result.
Verify that interval durations sum to the measurement period and that the final exposure is expressed in the requested units.

Why the Interval Sum Works

The interval formula follows from adding exposure over time. In a short period of $n_j$ days with $q_j$ covered units, there are $q_j n_j$ unit-days of exposure. Dividing by 365 converts unit-days into unit-years. Because the intervals partition the term, every day is included once and only once. Adding the unit-days across intervals therefore gives the total:

E_{\text{units}}=\frac{\sum_j q_j n_j}{365} =\sum_j q_j\frac{n_j}{365}.

For amount-based exposure, an amount $A_j$ represents $A_j/K$ units of the chosen amount base. Its contribution over $n_j$ days is $(A_j/K)n_j/365$ amount-unit-years. Adding these contributions gives the amount-based formula. This calculation measures exposure; it does not by itself calculate premium or establish that expected losses are exactly proportional to exposure.

Worked Example: Adding a Covered Vehicle Midterm

A hypothetical 365-day auto policy initially covers one vehicle. On day 121, an endorsement adds a second vehicle, which remains covered through the end of the term. Calculate exposure in vehicle-years, using a 365-day year.

Method: partition the term by the endorsement’s effective date. The first interval is 120 days, days 1 through 120, with one covered vehicle. The second is the remaining $365-120=245$ days, with two covered vehicles.

First interval: $1\times120=120$ vehicle-days.

Second interval: $2\times245=490$ vehicle-days.

Total exposure is $120+490=610$ vehicle-days. Converting to vehicle-years gives $610/365=1.6712$ vehicle-years, rounded to four decimals.

Check: $120+245=365$ days, so the intervals cover the whole term once. The result is greater than one vehicle-year, as expected, because the second vehicle is covered for part of the term.

Common mistake avoided: Counting one full vehicle-year for the original vehicle and one full vehicle-year for the added vehicle would give 2.0000 vehicle-years. That incorrectly gives the added vehicle exposure before its effective date.

Amount-Based Exposure

When the relevant base is insured amount, each interval contributes in proportion to the amount in force. It is often helpful to convert the amount to multiples of $K$ before multiplying by days. That keeps the units clear: an amount of $40,000 is four units if $K$ is $10,000.

Worked Example: An Insured Amount Changes Twice

A hypothetical property policy has a 365-day term. The insured amount is $40,000 for 100 days, $70,000 for the next 165 days, and $50,000 for the final 100 days. Calculate exposure in $10,000-of-insured-amount-years. Round to four decimals.

Method: normalize each amount and weight it by its interval. The three amounts correspond to $40{,}000/10{,}000=4$, $70{,}000/10{,}000=7$, and $50{,}000/10{,}000=5$ exposure units.

The interval contributions in normalized amount-unit-days are:

First interval: $4\times100=400$. Second interval: $7\times165=1{,}155$. Third interval: $5\times100=500$.

Their sum is $400+1{,}155+500=2{,}055$ normalized amount-unit-days. Thus exposure is $2{,}055/365=5.6301$ $10,000-of-insured-amount-years, rounded to four decimals.

Check: $100+165+100=365$ days. The weighted average amount is $5.6301\times$10,000$, which is between the lowest and highest amounts in force, as it should be.

Common mistake avoided: The weighted-day total is 2,055, not 2,555. Adding the interval products separately is a useful arithmetic check before dividing by 365.

Reconciling Overlapping Records

Policy data may include a declaration showing an original amount for the full term and endorsement records showing later changes. Those records can overlap in their stated date ranges because one describes the policy as issued and another describes a modification. They are not automatically separate blocks of exposure to add together. Reconstruct the amount actually in force for each interval, then calculate exposure from that timeline.

Worked Example: Avoiding Double Counting in an Endorsement Record

A hypothetical 365-day property policy was issued with a $60,000 insured amount. A midterm endorsement changes the amount to $90,000 beginning on day 101. A later endorsement changes it to $75,000 beginning on day 251. The policy record still displays the original $60,000 declaration for the full term. Calculate exposure in $10,000-of-insured-amount-years, to four decimals.

Method: use the final in-force amount on each successive interval. The amount is $60,000 for 100 days, $90,000 for 150 days (days 101 through 250), and $75,000 for the final $365-250=115$ days.

In units of $10,000, the respective amounts are 6, 9, and 7.5. Their interval contributions are $6\times100=600$, $9\times150=1{,}350$, and $7.5\times115=862.5$ normalized amount-unit-days.

The total is $600+1{,}350+862.5=2{,}812.5$. Divide by 365: $2{,}812.5/365=7.7055$ $10,000-of-insured-amount-years, rounded to four decimals.

Check: $100+150+115=365$ days, and the timeline uses the amount in force throughout the term.

Common mistake avoided: Adding the original $60,000 for a full year to the endorsement’s $90,000 for 150 days would produce $6+9(150/365)=9.6986$ exposure units. That counts the original amount during days when the active amount is different and adds the endorsement as if it were extra coverage. The corrected timeline replaces the original amount on the days affected by each endorsement.

Checks That Prevent Exposure Errors

A reliable calculation has both a timeline check and a units check. First confirm that all interval lengths sum to the period being measured. If they do not, there is either a gap, an overlap, or an incorrect day count. Next confirm that the amount or unit count on each interval is the quantity actually in force, not merely the value on an original policy snapshot.

Use the effective date. An endorsement’s processing or entry date is not necessarily its effective date. Use the effective date specified in the problem.
Do not add successive values as if they were simultaneous. If an amount changes from one level to another, the levels apply to different intervals unless the coverage explicitly says otherwise.
Do not multiply every recorded amount by the full term. Apply each amount only to the interval when it was in force.
Keep the exposure base visible. State whether the answer is vehicle-years, policy-years, or amount-based exposure, and state the amount unit when applicable.
Use the instructed time denominator. Do not silently change a stated 365-day convention or another stipulated basis.

Key takeaway: Midterm changes are handled by partitioning the measurement period into successive intervals and assigning each interval the exposure quantity actually in force. Sum those interval contributions once, using the specified exposure base and time denominator.

Check Your Understanding

For each problem, build the timeline before calculating. Use the stated time basis, and round exposure answers to four decimals where needed.

A hypothetical 365-day fleet policy covers three vehicles for the first 90 days. An endorsement adds one vehicle for the rest of the term. Find exposure in vehicle-years. (Answer: $1{,}370/365=3.7534$ vehicle-years.)
A hypothetical 365-day property policy has $30,000 of insured amount for 120 days and $50,000 for the remaining 245 days. Find exposure in $10,000-of-insured-amount-years. (Answer: $1{,}585/365=4.3425$ $10,000-of-insured-amount-years.)
A hypothetical 365-day policy has an insured amount of $80,000 for 60 days, $100,000 for 180 days, and $40,000 for 125 days. Find exposure in $10,000-of-insured-amount-years. (Answer: $2{,}780/365=7.6164$ $10,000-of-insured-amount-years.)
A hypothetical 360-day contract covers two units for 60 days, three units for the next 120 days, and one unit for the remaining 180 days. Find exposure in unit-years using a 360-day year. (Answer: $(2(60)+3(120)+1(180))/360=1.8333$ unit-years.)
A policy record shows $120,000 for a hypothetical 365-day term, but an endorsement reduces the amount to $80,000 effective on day 181, leaving 185 days. Find exposure in $10,000-of-insured-amount-years and explain why the original record is not added for the entire term. (Answer: $(12(180)+8(185))/365=3{,}640/365=9.9726$ $10,000-of-insured-amount-years; the $80,000 amount replaces the original amount for the remaining interval.)

What You'll Learn