Applying Minimum and Maximum Modification Factors

Why Minimum and Maximum Factors Matter

An experience modification summarizes how a risk’s experience compares with the plan’s expected experience. The calculation may produce an uncapped modification outside the range that the rating plan permits. A plan can limit the factor that is ultimately used by setting a minimum, or floor, and a maximum, or ceiling. Applying these limits changes the rating factor; it does not change the underlying losses or the uncapped calculation.

This tutorial concerns the direct application of stated limits to a completed uncapped factor. Assume throughout that the plan specifies one floor and one ceiling for the same modification, that the floor is no greater than the ceiling, and that no other adjustment is to be applied between the uncapped calculation and these limits. The limits are plan terms, not universal values.

Definition: Let $M_0$ be the uncapped modification, $M_{\min}$ the stated minimum factor, and $M_{\max}$ the stated maximum factor, with $M_{\min}\le M_{\max}$. The permitted modification is $$ M=\min\left(M_{\max},\max\left(M_{\min},M_0\right)\right). $$ Thus, a raw result below the floor is raised to the floor, a result within the allowed range is retained, and a result above the ceiling is reduced to the ceiling.

A modification factor is dimensionless. It is not a dollar amount, and a floor or ceiling on the factor is not a floor or ceiling on the risk’s losses. For example, a floor of 0.90 means that the modification used cannot be less than 0.90; it does not mean that losses are increased by 0.90 dollars. If a premium calculation is required, the permitted factor—not the uncapped factor—is the factor to carry forward under the stated plan.

The Limit Operation

The calculation has two comparisons. First compare the uncapped factor with the minimum. The inner operation, $\max(M_{\min},M_0)$, ensures the value is at least the floor. Then compare that result with the maximum. The outer operation, $\min(M_{\max},\cdot)$, ensures the value is no greater than the ceiling.

The nesting is intentional. If the raw factor is below the floor, the first comparison raises it to the floor, and the second comparison leaves it there because the floor is no greater than the ceiling. If the raw factor is within the range, the first comparison retains it and the second comparison also retains it. If the raw factor is above the ceiling, the first comparison retains it, and the second comparison lowers it to the ceiling.

Calculate the uncapped factor.
Complete the experience-modification calculation specified by the plan and label its result $M_0$. Do not treat a limit as an input to that calculation unless the plan explicitly says to do so.

Check the stated range.
Record $M_{\min}$ and $M_{\max}$, and verify that $M_{\min}\le M_{\max}$. Both limits must be expressed as factors on the same scale as $M_0$.

Apply the limits and verify.
Use $M=\min(M_{\max},\max(M_{\min},M_0))$. Confirm that the result satisfies $M_{\min}\le M\le M_{\max}$.

Worked Applications

The examples use hypothetical plans that explicitly provide their uncapped calculation rule. That rule is included only to make each calculation traceable; actual plans may use different formulas. In each example, amounts in the uncapped formula are dollars, while the resulting modification and its limits are dimensionless factors. No interest rate, mortality assumption, or probability distribution is used: each calculation is deterministic from the stated plan inputs.

Worked Example: Raising a Result Below the Floor

A hypothetical plan calculates the uncapped modification as $M_0=1+Z(A-E)/(E+B)$, where $A$ is actual loss, $E$ is expected loss, $B$ is a stabilizing amount, and $Z$ is a credibility weight. For a risk, $A=\$72{,}000$, $E=\$100{,}000$, $B=\$40{,}000$, and $Z=0.60$. The plan sets $M_{\min}=0.90$ and $M_{\max}=1.20$. Find the permitted modification.

Method. Calculate the uncapped factor using the plan’s stated formula, then apply both limits to that completed result.

Step 1: Calculate the loss deviation. $$ A-E=\$72{,}000-\$100{,}000=-\$28{,}000. $$

Step 2: Calculate the denominator. $$ E+B=\$100{,}000+\$40{,}000=\$140{,}000. $$

Step 3: Calculate the adjustment and uncapped factor. $$ \frac{Z(A-E)}{E+B} =\frac{0.60(-\$28{,}000)}{\$140{,}000} =\frac{-\$16{,}800}{\$140{,}000} =-0.12, \qquad M_0=1-0.12=0.88. $$

Step 4: Apply the floor and ceiling. Since $0.88<0.90$, the result is below the floor. The floor raises it to 0.90. The ceiling does not reduce it, because $0.90<1.20$. $$ M=\min(1.20,\max(0.90,0.88)) =\min(1.20,0.90) =0.90. $$

Check. The final factor satisfies $0.90\le0.90\le1.20$. The correct permitted modification is 0.90. The common mistake is to report 0.88 because it is the raw result; the plan’s floor means that 0.88 is not the factor permitted for use.

Worked Example: Retaining a Result Within the Range

A hypothetical plan uses $M_0=1+Z(A-E)/(E+B)$. A small equipment-maintenance account has $A=\$118{,}000$, $E=\$100{,}000$, $B=\$50{,}000$, and $Z=0.50$. Its modification limits are 0.85 and 1.15. Find the permitted modification. All dollar amounts are deterministic plan inputs; no interest, mortality, or probability distribution is assumed.

Step 1: Find the deviation. $$ A-E=\$118{,}000-\$100{,}000=\$18{,}000. $$

Step 2: Find the denominator. $$ E+B=\$100{,}000+\$50{,}000=\$150{,}000. $$

Step 3: Calculate the adjustment and raw result. $$ \frac{Z(A-E)}{E+B} =\frac{0.50(\$18{,}000)}{\$150{,}000} =\frac{\$9{,}000}{\$150{,}000} =0.06, \qquad M_0=1+0.06=1.06. $$

Step 4: Compare with both limits. The raw result is above the floor and below the ceiling: $0.85<1.06<1.15$. Neither limit changes it. $$ M=\min(1.15,\max(0.85,1.06)) =\min(1.15,1.06) =1.06. $$

Check. The final factor remains within the stated range. The permitted modification is 1.06. The common mistake is to assume that every factor must be moved toward 1.00. The limits do not do that: they change a factor only when it falls outside the plan’s permitted range.

Worked Example: Reducing a Result Above the Ceiling

A hypothetical plan again specifies $M_0=1+Z(A-E)/(E+B)$. A warehouse has $A=\$190{,}000$, $E=\$100{,}000$, $B=\$50{,}000$, and $Z=0.80$. Its minimum factor is 0.90 and its maximum factor is 1.25. Determine the permitted modification. No interest rate, mortality assumption, or probability distribution is applicable to these stipulated inputs.

Step 1: Calculate the deviation. $$ A-E=\$190{,}000-\$100{,}000=\$90{,}000. $$

Step 2: Calculate the denominator. $$ E+B=\$100{,}000+\$50{,}000=\$150{,}000. $$

Step 3: Calculate the uncapped adjustment. $$ \frac{Z(A-E)}{E+B} =\frac{0.80(\$90{,}000)}{\$150{,}000} =\frac{\$72{,}000}{\$150{,}000} =0.48, \qquad M_0=1+0.48=1.48. $$

Step 4: Apply the ceiling. Since $1.48>1.25$, the raw result exceeds the maximum. The floor has no effect because the raw result is already above it. $$ M=\min(1.25,\max(0.90,1.48)) =\min(1.25,1.48) =1.25. $$

Check. The final factor satisfies $0.90\le1.25\le1.25$. The permitted modification is 1.25. The common mistake is to stop after observing that the raw factor exceeds the minimum. Meeting the floor does not remove the separate requirement to check the ceiling.

Why the Formula Works

The nested minimum and maximum expression follows directly from the definition of the permitted interval. First, $\max(M_{\min},M_0)$ is at least $M_{\min}$: it equals the floor if $M_0$ is lower, and equals $M_0$ otherwise. Next, taking the minimum with $M_{\max}$ ensures the final value is no greater than $M_{\max}$. Because $M_{\min}\le M_{\max}$, this second comparison cannot push the result below the floor.

The formula also retains every raw factor already within the interval. If $M_{\min}\le M_0\le M_{\max}$, then $\max(M_{\min},M_0)=M_0$, and $\min(M_{\max},M_0)=M_0$. Thus the limits alter only a result outside the allowed range, and the resulting factor always satisfies both bounds.

This operation is appropriate when both stated limits apply directly to the same completed uncapped factor. It does not determine how a plan defines the uncapped calculation, whether a limit applies to a particular coverage, or whether other adjustments are required. Those details must come from the plan’s stated rules. If the floor exceeds the ceiling, the inputs are inconsistent with this setup; do not silently choose one limit or reverse them.

Common Mistakes and Practical Checks

Applying the wrong bound. A floor raises a result that is too low; a ceiling lowers a result that is too high. Identify which comparison fails before changing the raw value.
Reversing the nested operations. The correct expression is $\min(M_{\max},\max(M_{\min},M_0))$. Reversing the roles of the minimum and maximum can produce a value outside the permitted range.
Capping an input instead of the completed factor. Unless the plan explicitly requires otherwise, first calculate $M_0$ from the prescribed formula. The floor and ceiling in this tutorial are applied to $M_0$, not separately to actual losses, expected losses, or the credibility weight.
Assuming the limits are symmetric around 1.00. A plan’s stated limits need not be equally distant from 1.00. Use the actual values given rather than inventing a corresponding limit.
Failing to verify the result. After applying the limits, check $M_{\min}\le M\le M_{\max}$. If this inequality fails, revisit the comparisons and confirm that the bounds were recorded in the correct order.

Key takeaway: Calculate the uncapped modification first, then apply the plan’s stated floor and ceiling to that result. The permitted factor is unchanged when the raw result is within the range, raised to the floor when it is below, and reduced to the ceiling when it is above.

Check Your Understanding

For each question, assume the stated minimum and maximum apply directly to the completed uncapped factor. No interest, mortality, or probability distribution is used unless expressly stated.

A hypothetical plan produces an uncapped factor of 0.82 and specifies a floor of 0.88 and a ceiling of 1.18. What factor is permitted? (Answer: 0.88)
A plan produces an uncapped factor of 1.11, with a floor of 0.90 and a ceiling of 1.10. Which limit applies, and what is the final factor? (Answer: The ceiling applies; 1.10)
A plan’s raw result is $M_0=1+Z(A-E)/(E+B)$, with $A=\$84{,}000$, $E=\$100{,}000$, $B=\$60{,}000$, and $Z=0.50$. If the floor is 0.92 and the ceiling is 1.12, calculate the permitted factor. (Answer: $M_0=1-0.05=0.95$, so the permitted factor is 0.95)
A hypothetical plan uses the same stated formula, with $A=\$160{,}000$, $E=\$100{,}000$, $B=\$50{,}000$, and $Z=0.50$. The floor is 0.90 and the ceiling is 1.18. Calculate the uncapped and permitted factors. (Answer: $M_0=1+0.20=1.20$; the permitted factor is 1.18)
A plan calculation produces $M_0=1.04$, but the stated minimum is 1.06 and the stated maximum is 1.03. Can the direct floor-and-ceiling rule be applied as defined here? (Answer: No. The stated bounds are inconsistent because $1.06>1.03$; clarify the plan terms before selecting a factor.)

What You'll Learn