Why Experience-Rating Plans Need Boundaries
An experience-rating plan can indicate different adjustments for risks with different loss experience. But an indicated adjustment does not have to flow without limit into the premium charged. A plan may set a minimum or maximum adjustment factor, a minimum or maximum premium, or a limit on the change from the expiring premium. These boundaries define the permitted range of plan outcomes; they do not determine the underlying experience modification.
The design question is therefore not simply “What adjustment does the experience indicate?” It is also “Which parts of the premium may change, and how far may the final premium move?” A boundary can constrain the adjustment itself, the dollars resulting from the adjustment, or the change between policy periods. Those are different choices and can produce different results.
Let \(N\) be the portion of the prospective premium base that the plan does not adjust, and let \(A\) be the portion that it does adjust. Let \(m\) denote an indicated adjustment factor, supplied by the rating process. The factor-bound version of the premium calculation is:
where \(L\) and \(U\) are the factor minimum and maximum, with \(L\leq U\). This setup keeps the adjustable portion distinct from the nonadjustable portion. It is not a method for calculating \(m\); it shows how stated boundaries constrain a supplied factor.
What a Factor Boundary Does
A lower factor limit prevents the factor applied to the adjustable premium component from falling below a stated value. An upper factor limit prevents it from exceeding another stated value. For example, a plan might allow factors only from 0.75 to 1.25. The range limits the adjustment to the adjustable component; it does not automatically place the total premium between 75% and 125% of the prospective premium base.
The distinction matters whenever some premium is nonadjustable. If a factor is applied only to \(A\), the premium is \(N+A m_b\), not \((N+A)m_b\). The latter calculation also changes \(N\), contradicting the stated plan design.
Separate the nonadjustable component \(N\) from the adjustable component \(A\), using the plan’s stated definitions.
Compare the indicated factor \(m\) with \(L\) and \(U\). Use \(L\) if \(m<L\), \(m\) if \(L\leq m\leq U\), and \(U\) if \(m>U\).
Apply the bounded factor to \(A\), then add \(N\). Check whether any separately stated premium-dollar boundaries also apply.
The clipping formula follows directly from the three possible cases. If \(m<L\), then \(\max(L,m)=L\), and because \(L\leq U\), the outer minimum leaves \(L\). If \(L\leq m\leq U\), the inner maximum is \(m\), and the outer minimum also leaves \(m\). If \(m>U\), the inner maximum is \(m\), and the outer minimum gives \(U\). Thus the result always lies in \([L,U]\), while every factor already in that range is unchanged.
Worked Example: Bounding Only the Adjustable Component
A hypothetical commercial policy has a prospective premium base of $6,000, consisting of $1,800 that is nonadjustable and $4,200 that is adjustable. The indicated factor is 1.38. The plan’s factor range is 0.75 to 1.25. Find the premium after applying the factor boundary, and compare it with the result without that boundary.
Method. Bound the factor first, then apply it only to the adjustable component. The raw factor exceeds the maximum:
The adjusted component is \(4{,}200(1.25)=\$5{,}250\). Adding the unchanged component gives a premium of \(\$1{,}800+\$5{,}250=\$7{,}050\). Without the factor boundary, the premium would be \(\$1{,}800+4{,}200(1.38)=\$1{,}800+\$5{,}796=\$7{,}596\). The boundary reduces this result by \(\$7{,}596-\$7{,}050=\$546\). The bounded premium is above the $6,000 base, as expected for a bounded factor of 1.25 applied to the adjustable portion.
Common mistake avoided: Applying 1.25 to the entire $6,000 base gives $7,500. That is wrong because the plan specifies that only $4,200 is adjustable; the correct result is $7,050.
Dollar Boundaries and Renewal-Change Boundaries
A dollar minimum or maximum constrains the premium amount rather than the factor. For a premium \(P\) after the factor boundary, a dollar floor \(P_{\min}\) and ceiling \(P_{\max}\) can be represented as:
A dollar boundary may bind even when the factor is within its allowed range. Conversely, the factor boundary may bind while the dollar boundary does not. A factor range is often easier to describe as a limit on experience adjustment; dollar bounds are direct constraints on the premium charged. Neither one implies the other without knowing the premium base and the plan’s adjustable components.
A renewal-change limit instead compares the premium with a specified expiring premium \(P_{\text{prior}}\). If the permitted decrease is at most \(r_d\) and the permitted increase is at most \(r_u\), with both rates expressed as nonnegative proportions, the allowed renewal interval is:
This is a limit relative to the prior premium, not necessarily relative to the current prospective premium base. It can therefore behave differently from a factor cap or a fixed-dollar ceiling. When more than one boundary applies, the plan must specify their order and the reference amount for each one. Applying an annual change limit to a different premium stage can produce a different result.
Worked Example: Combining Factor and Dollar Limits
A hypothetical policy has a $2,400 nonadjustable component and a $6,000 adjustable component. The factor range is 0.70 to 1.30, the dollar minimum is $6,800, and the dollar maximum is $10,000. Apply the boundaries to an indicated factor of 0.50 and, separately, to an indicated factor of 1.55. The plan specifies that the factor boundary is applied first and the dollar boundary second.
Low indicated factor. Since \(0.50<0.70\), the bounded factor is 0.70. The premium before dollar limits is \(\$2{,}400+\$6{,}000(0.70)=\$2{,}400+\$4{,}200=\$6{,}600\). This is below the $6,800 floor and below the $10,000 ceiling, so the final premium is $6,800.
High indicated factor. Since \(1.55>1.30\), the bounded factor is 1.30. The premium before dollar limits is \(\$2{,}400+\$6{,}000(1.30)=\$2{,}400+\$7{,}800=\$10{,}200\). This exceeds the $10,000 ceiling, so the final premium is $10,000.
Check. The two final premiums, $6,800 and $10,000, fall within the stated dollar range. The floor binds for the low case; the ceiling binds for the high case. The factor limits bind in both cases. The example illustrates that factor limits and dollar limits can both affect an outcome, and it follows the specified sequence rather than assuming an unstated order.
Common mistake avoided: Stopping after calculating $6,600 or $10,200 ignores the separately stated dollar bounds. The factor limits do not replace the dollar floor or ceiling.
Choose Boundaries for a Defined Purpose
A boundary is useful only if its role is clear. A factor minimum or maximum constrains the permitted adjustment to the adjustable component. A premium floor or ceiling constrains the amount charged. A renewal-change cap constrains movement relative to the expiring premium. These limits may overlap, but they are not interchangeable.
| Boundary | What it constrains | Reference point | Design question |
|---|---|---|---|
| Factor minimum or maximum | The adjustment factor | Stated factor range | How far may the factor depart from its permitted range? |
| Dollar premium minimum or maximum | The resulting premium in dollars | Stated dollar amount | What premium amount is the plan permitted to produce? |
| Renewal-change limit | The change between policy periods | Specified expiring premium | What percentage change from the prior premium is allowed? |
Before applying any combined limits, state the order. For example, a plan might calculate the premium using a bounded factor, apply a dollar floor and ceiling, and then apply a renewal-change limit. Another plan could specify a different sequence. Because each stage can alter the amount passed to the next one, an analyst should not silently rearrange the steps or assume that the final result is unchanged.
It is also important to state which premium amount serves as the base for a limit. A percentage cap relative to the expiring premium is not the same as a percentage cap relative to the prospective premium base. A dollar minimum on the final policy premium is not the same as a minimum factor on one component. Clear definitions prevent a technically correct calculation from implementing the wrong plan.
Worked Example: Applying a Renewal-Change Limit
A hypothetical account had an expiring premium of $10,000. After applying the plan’s factor and dollar boundaries, its renewal premium is $13,200. The plan limits the increase from the expiring premium to 15% and the decrease to 10%. A separate minimum-premium provision requires at least $8,500. Find the final premium for this case and for a second case whose premium after factor and dollar boundaries is $8,800. The plan applies the renewal-change limit after the other stated premium boundaries.
Step 1: Calculate the renewal interval. The lower end is \(\$10{,}000(1-0.10)=\$9{,}000\). The upper end is \(\$10{,}000(1+0.15)=\$11{,}500\). Therefore, the renewal-change boundary permits premiums from $9,000 through $11,500.
Step 2: Apply it to the $13,200 case. Since $13,200 exceeds $11,500, the final premium is $11,500. The premium is $1,700 below the otherwise calculated renewal amount and is exactly 15% above the expiring premium.
Step 3: Apply it to the $8,800 case. Since $8,800 is below $9,000, the renewal-change boundary raises it to $9,000. The separate $8,500 minimum does not bind because $9,000 is already greater than $8,500.
Check. Both final amounts lie in the allowed interval \([\$9{,}000,\$11{,}500]\), and both satisfy the $8,500 minimum. The calculation uses the expiring premium—not the prospective premium base—as the reference for the percentage limits.
Common mistake avoided: Applying 15% to the $13,200 renewal indication would give $15,180, which is not the specified limit. The permitted increase is measured from the $10,000 expiring premium, so the correct upper bound is $11,500.
Common Design and Calculation Traps
- Applying a factor to the wrong premium base. Apply it to the component identified as adjustable, then add any nonadjustable component.
- Confusing a factor limit with a dollar limit. A factor can be within range while the resulting premium still violates a separate dollar floor or ceiling.
- Using the wrong reference premium. A renewal-change cap must use the reference specified by the plan, commonly the expiring premium, rather than an assumed base.
- Leaving the order unstated. If multiple limits apply, specify which premium enters each stage and in what order.
- Treating a boundary as the experience calculation. A boundary constrains an indicated adjustment; it does not explain how that indication was produced.
Careful boundary design makes a plan’s permitted outcomes explicit and supports consistent application across risks. This treatment is aligned with the CAS Exam 8 (Advanced Ratemaking) syllabus.
Check Your Understanding
For each problem, identify which boundary binds and check the final amount against the stated limits.
- A $5,500 base consists of $1,000 nonadjustable and $4,500 adjustable premium. The indicated factor is 1.40, with factor limits of 0.80 and 1.20. What is the premium after the factor boundary? (Answer: $6,400.)
- A $900 nonadjustable component and a $3,100 adjustable component are subject to factor limits of 0.75 and 1.25, then a $3,500 dollar floor and $4,800 ceiling. If the indicated factor is 0.60, what is the final premium? (Answer: $3,500.)
- A policy’s expiring premium is $18,000. Its indicated renewal premium after other stated boundaries is $21,240, and the plan allows an increase of at most 12% from expiring premium. What premium results from the renewal-change limit? (Answer: $20,160.)
- A plan applies a factor range of 0.70 to 1.30, then dollar limits of $7,000 to $12,000. The nonadjustable premium is $2,000, the adjustable premium is $8,000, and the indicated factor is 1.40. What is the final premium, and which limits bind? (Answer: $12,000; the factor maximum and dollar ceiling bind.)
- A prior premium is $15,000. The premium after factor and dollar limits is $12,900. A renewal-change boundary allows no decrease greater than 10%, and a separate minimum premium is $13,200. The renewal-change boundary is applied last. What is the final premium? (Answer: $13,500; the renewal-change lower bound binds and the separate minimum is satisfied.)