Quantify Premium Range Under Alternative Parameters

What a Premium Range Measures

A retrospective premium depends on the loss amount included in the plan calculation, but it is also subject to contractual parameters. A loss conversion factor changes the amount of loss reflected in the premium; a loss cap limits the loss amount used in the calculation; an expense provision contributes a fixed amount; and minimum and maximum premiums bound the final charge. Comparing alternatives requires applying those pieces in the right order and comparing both premium endpoints.

The premium range is the difference between the highest and lowest premium attainable over the stated range of losses. It is not necessarily the difference between the maximum premium and the minimum premium named in the contract: the modeled premiums may never reach one or both bounds. Nor is it necessarily the difference between the uncapped loss-based premium at two loss levels. The calculation must apply the loss cap and then the premium bounds.

Definition: Let $X$ be the loss amount used in the retrospective calculation, $L\geq 0$ the loss conversion factor, $C\geq 0$ the loss cap, $E$ the fixed expense provision, and $m\leq M$ the minimum and maximum premiums. For a plan that caps losses before applying the conversion factor, the premium is $$ P(X)=\min\!\left(M,\max\!\left(m,E+L\min(X,C)\right)\right). $$ The premium range over an allowable loss interval is the largest attainable value of $P(X)$ minus its smallest attainable value on that interval.

This formula assumes that the expense provision is a fixed dollar amount included in the premium subject to the minimum and maximum, that $L$ is nonnegative, and that the loss cap $C$ applies to $X$ before conversion. A plan that caps the converted loss component instead has a different formula; do not substitute one cap convention for another. This tutorial treats the parameters as deterministic contractual amounts. No interest, mortality assumption, or probability distribution is needed to calculate a contractual range.

Endpoint Formula and Justification

Because $L\geq 0$, increasing $X$ cannot decrease $\min(X,C)$. Multiplication by $L$, addition of $E$, and application of the minimum and maximum premium bounds preserve that nondecreasing relationship. Therefore, over an interval $x_{\min}\leq X\leq x_{\max}$, the smallest premium occurs at $x_{\min}$ and the largest at $x_{\max}$.

Write the bounding operation as $\operatorname{clip}_{m,M}(z)=\min(M,\max(m,z))$. Substituting the two loss endpoints into the plan formula gives the range width:

R= \operatorname{clip}_{m,M}\!\left(E+L\min(x_{\max},C)\right) - \operatorname{clip}_{m,M}\!\left(E+L\min(x_{\min},C)\right).

For the common case $0\leq X$ with no stated upper limit on losses, $x_{\min}=0$, and the cap makes the largest loss-based amount attainable at $X\geq C$. Thus:

P_{\mathrm{low}}=\min(M,\max(m,E)),\qquad P_{\mathrm{high}}=\min(M,\max(m,E+LC)),\qquad R=P_{\mathrm{high}}-P_{\mathrm{low}}.

This endpoint argument is the key justification: every intermediate loss produces a premium between the premiums at the endpoint losses. If $L=0$, the premium does not vary with losses and the range width is zero. If $C=0$, there is no loss component in this formula, so the premium is also constant over losses.

Fix the loss domain and cap convention.
Identify the smallest and largest possible $X$, and determine whether the stated cap applies before or after conversion.

Calculate the uncapped premium at each loss endpoint.
Apply the loss cap to each endpoint, multiply the capped loss by $L$, and add $E$.

Apply the premium bounds separately.
Raise any endpoint below $m$ to $m$, and lower any endpoint above $M$ to $M$.

Subtract the bounded endpoints.
The upper premium minus the lower premium is the attainable range width.

Worked Examples

Worked Example: Comparing Conversion Factors and Loss Caps

A hypothetical plan has a minimum premium of $30,000, a maximum premium of $100,000, and an expense provision of $12,000. Losses range from $0 upward, and the loss cap is applied before conversion. Compare: Plan A, $L=1.20$ and $C=\$50,000$; Plan B, $L=1.50$ and $C=\$50,000$; and Plan C, $L=1.50$ and $C=\$40,000$. Calculate each bounded premium range and identify how the parameter changes affect it. These are deterministic hypothetical amounts; no interest, mortality assumption, or loss distribution is specified or required.

Method. For each plan, evaluate $E$ at zero loss and $E+LC$ at a loss at or above the cap. Then apply the $30,000 floor and $100,000 maximum. This avoids comparing unbounded amounts when one endpoint may be clipped.

Plan A. At $X=0$, the uncapped premium is $12,000+1.20(0)=\$12,000$, which is below the $30,000 minimum. Therefore, the lower premium is $30,000. At $X\geq \$50,000$, the capped loss is $50,000 and the uncapped premium is $12,000+1.20(50,000)=12,000+60,000=\$72,000$. This is between the bounds, so the upper premium is $72,000. The range width is $72,000-30,000=\$42,000$.

Plan B. The zero-loss premium is again $12,000, so the lower premium remains $30,000. At the cap, the uncapped premium is $12,000+1.50(50,000)=12,000+75,000=\$87,000$. It is below the maximum, so the upper premium is $87,000. The range width is $87,000-30,000=\$57,000$.

Plan C. The zero-loss premium is $12,000, giving a bounded lower premium of $30,000. At the $40,000 cap, the uncapped premium is $12,000+1.50(40,000)=12,000+60,000=\$72,000$. The upper premium is $72,000, and the range width is $72,000-30,000=\$42,000$.

Comparison and check. Raising $L$ from 1.20 to 1.50 while holding $C$ fixed increases the upper endpoint by $15,000 and leaves the lower endpoint unchanged; Plan B’s range is therefore $15,000 wider than Plan A’s. Reducing $C$ from $50,000 to $40,000 in Plan C reduces the upper endpoint by $15,000 relative to Plan B, returning the width to $42,000. In each plan the upper endpoint exceeds the lower endpoint, and each premium endpoint respects its stated bound.

Common mistake avoided: Do not apply the loss cap to the converted amount in this example. The stated convention is $L\min(X,C)$, so the cap is applied to $X$ first.

Worked Example: Expense Changes When the Maximum Premium Binds

Two hypothetical alternatives have $L=1.00$, $C=\$70,000$, minimum premium $m=\$20,000$, and maximum premium $M=\$85,000$. Alternative A has $E=\$8,000$; Alternative B has $E=\$24,000$. Losses range from zero upward, and the cap applies before conversion. Find each premium range and explain the change. No interest, mortality assumption, or loss distribution is needed.

Method. Calculate both raw endpoint premiums for each alternative, then apply the minimum and maximum separately. In particular, an expense increase can move both endpoints, while the maximum premium may prevent the upper endpoint from increasing by the full expense change.

Alternative A. At zero loss, the raw premium is $E=\$8,000$, below the $20,000 minimum, so the bounded lower endpoint is $20,000. At the loss cap, the raw premium is $8,000+1.00(70,000)=\$78,000$, below the $85,000 maximum. The bounded upper endpoint is $78,000. The range width is $78,000-20,000=\$58,000$.

Alternative B. At zero loss, the raw premium is $24,000, which exceeds the $20,000 minimum, so the lower endpoint is $24,000. At the loss cap, the raw premium is $24,000+1.00(70,000)=\$94,000$. The maximum premium clips this to $85,000. The bounded range width is $85,000-24,000=\$61,000$.

Comparison and check. The upper endpoint rises from $78,000 in A to the $85,000 cap in B, an increase of $7,000. The lower endpoint rises from $20,000 to $24,000, an increase of $4,000. Therefore, the range increases by $7,000-4,000=\$3,000$, from $58,000 to $61,000. The endpoints also verify the widths directly: $78,000-20,000=58,000$ and $85,000-24,000=61,000$.

Before applying the premium bounds, subtracting Alternative A’s minimum premium from Alternative B’s raw upper amount gives $94,000-20,000=\$74,000$. That $74,000 is not Alternative B’s actual range: it combines B’s raw upper endpoint with a lower endpoint that is not attainable under B, and it ignores the maximum premium. The correct B range uses its own bounded endpoints: $85,000-24,000=\$61,000$.

Common mistake avoided: Do not subtract the contractual minimum from an uncapped upper amount and call that the actual range. First determine which bounds bind for the same alternative.

Worked Example: Range Over a Restricted Loss Interval

A hypothetical plan is evaluated only for losses from $10,000 through $90,000. Its minimum premium is $25,000, maximum premium is $90,000, expense provision is $18,000, and loss cap is $60,000, applied before conversion. Compare Plan A, $L=1.10$, with Plan B, $L=1.40$. Calculate the premium ranges over the stated loss interval. Assume no interest, mortality, or probability distribution; the interval alone defines the loss scenarios.

Method. Since the interval does not begin at zero, use $x_{\min}=\$10,000$ and $x_{\max}=\$90,000$ in the general endpoint formula. The upper endpoint is evaluated at the capped loss, but the lower endpoint is not a zero-loss premium.

Plan A lower endpoint. The loss at the lower endpoint is below the $60,000 cap. Its raw premium is $18,000+1.10(10,000)=18,000+11,000=\$29,000$. This exceeds the $25,000 minimum, so the bounded lower endpoint is $29,000.

Plan A upper endpoint. At $90,000 of losses, the capped amount is $60,000. The raw premium is $18,000+1.10(60,000)=18,000+66,000=\$84,000$. It is below the $90,000 maximum. The bounded upper endpoint is $84,000, and the range width is $84,000-29,000=\$55,000$.

Plan B lower endpoint. At $10,000 of losses, the raw premium is $18,000+1.40(10,000)=18,000+14,000=\$32,000$. It is above the minimum, so the bounded lower endpoint is $32,000.

Plan B upper endpoint. At $90,000 of losses, the capped amount is $60,000. The raw premium is $18,000+1.40(60,000)=18,000+84,000=\$102,000$. The maximum premium clips it to $90,000. The range width is $90,000-32,000=\$58,000$.

Comparison and check. The conversion factor increase raises the lower endpoint by $3,000 and raises the upper endpoint by $6,000 after clipping. The range therefore widens by $3,000, from $55,000 to $58,000. Both calculations use the stated loss interval, and the upper endpoint for B respects the maximum premium.

Common mistake avoided: Using zero loss as the lower endpoint would incorrectly apply the minimum premium here. The plan is being compared only for losses of at least $10,000.

How to Interpret Parameter Changes

The premium range is affected by more than the difference between the loss cap and zero. Its width depends on which endpoints are moved by a parameter change and whether either endpoint is clipped by a premium bound.

Parameter change	Typical effect before premium bounds	Important qualification
Increase $L$	Increases the loss-based premium at positive capped losses	The range may not increase if the affected endpoint is already at the maximum, or if both endpoints move equally.
Increase $C$	Raises the maximum loss amount included in the calculation when the cap is reached	The upper premium may already be at $M$, so a higher cap can leave the bounded range unchanged.
Increase $E$	Shifts raw premiums upward by the same dollar amount	The lower and upper endpoints can react differently when one crosses $m$ or reaches $M$.
Change $m$ or $M$	Changes the contractual premium bounds	A bound matters only when it clips an endpoint for the loss domain under consideration.

For a full loss domain beginning at zero, a higher expense provision does not automatically widen the range. If neither premium bound is active, it shifts both endpoints up by the same amount, leaving the width unchanged. If the lower endpoint is held at the minimum while the upper endpoint rises, the width can increase. If the upper endpoint is already at the maximum, a further expense increase may raise the lower endpoint but leave the upper endpoint fixed, causing the width to narrow. The binding bounds—not a general rule that expenses widen or narrow the range—determine the result.

Similarly, a higher $L$ or $C$ raises the raw upper endpoint when losses reach the cap, but it does not necessarily change the bounded upper premium. Once that endpoint has reached $M$, further increases in the raw amount do not increase the charged premium under this formula. When comparing alternatives, report the lower endpoint, upper endpoint, and width for each one; that makes the effect and any clipping visible.

Common Exam Traps

Mixing up the two caps. The loss cap $C$ limits the loss amount in the formula. The maximum premium $M$ limits the resulting premium. They are different parameters and may both bind.
Applying a floor or maximum only to the final range width. Bound each premium endpoint first, then subtract. A range width cannot be corrected reliably by clipping a raw difference.
Using the wrong loss interval. If losses are restricted to $[x_{\min},x_{\max}]$, evaluate the premium at those endpoints, not at zero and at the cap unless those are the stated endpoints.
Attributing a width change to only one endpoint. The change in width equals the change in the upper endpoint minus the change in the lower endpoint. For example, an upper endpoint increase of $7,000 and a lower endpoint increase of $4,000 widen the range by $3,000.
Comparing inconsistent alternatives. Keep the loss domain and cap convention clear. If one alternative uses a cap before conversion and another uses a cap after conversion, they are not directly comparable until the formulas are specified consistently.

Key takeaway: Calculate each alternative’s attainable lower and upper premium using its own loss interval, loss conversion factor, cap, expense provision, minimum, and maximum. Apply premium bounds to both endpoints before subtracting. The resulting width—and the endpoint changes that produce it—explains how the alternative parameters affect the premium range.

Check Your Understanding

For each problem, apply the stated cap convention and premium bounds before reporting the range width.

A hypothetical plan has $m=\$15,000$, $M=\$60,000$, $E=\$9,000$, $L=1.25$, and $C=\$32,000$. Losses range from zero upward, and the cap applies before conversion. What is the bounded premium range width? (Answer: $34,000.)
A plan has $m=\$25,000$, $M=\$100,000$, $E=\$30,000$, $L=0.80$, and $C=\$50,000$. Losses range from zero upward. Find the lower premium, upper premium, and range width. (Answer: $30,000; $70,000; $40,000.)
Two alternatives have $m=\$20,000$, $M=\$75,000$, $E=\$10,000$, and $C=\$40,000$, with losses ranging from zero upward. Alternative A has $L=1.00$; Alternative B has $L=1.50$. How much wider is B’s premium range? (Answer: $20,000 wider; A’s width is $30,000 and B’s is $50,000.)
A plan has $m=\$30,000$, $M=\$95,000$, $E=\$22,000$, $L=1.20$, and $C=\$50,000$. Losses range from zero upward. Calculate the bounded premium range width and state which endpoint is clipped. (Answer: $52,000; the lower endpoint is clipped to $30,000.)
A plan is evaluated only for losses from $5,000 to $40,000. It has $m=\$18,000$, $M=\$55,000$, $E=\$12,000$, $L=1.10$, and $C=\$30,000$, with the cap applied before conversion. Find the premium range width. (Answer: $27,000.)

Parameter change	Typical effect before premium bounds	Important qualification
Increase \(L\)	Increases the loss-based premium at positive capped losses	The range may not increase if the affected endpoint is already at the maximum, or if both endpoints move equally.
Increase \(C\)	Raises the maximum loss amount included in the calculation when the cap is reached	The upper premium may already be at \(M\), so a higher cap can leave the bounded range unchanged.
Increase \(E\)	Shifts raw premiums upward by the same dollar amount	The lower and upper endpoints can react differently when one crosses \(m\) or reaches \(M\).
Change \(m\) or \(M\)	Changes the contractual premium bounds	A bound matters only when it clips an endpoint for the loss domain under consideration.

What You'll Learn