Tutorials › CAS Exam 8 - Advanced Ratemaking › Applying Premium Caps And Floors After Credibility

CAS Exam 8 · Credibility in Experience Rating · Tutorial 484 of 989

Applying Premium Caps And Floors After Credibility

Calculate capped and floored experience-rated premiums, and identify when limits on the premium differ from limits on the experience modification.

Advanced 11 min read

What You'll Learn

  • Calculate an experience modification from a credibility-weighted experience ratio
  • Apply dollar premium floors and caps after credibility has been applied
  • Apply lower and upper bounds directly to an experience modification
  • Identify when dollar and modification limits are mathematically equivalent
  • Explain why limits on the modification can produce different premiums across risks

Why Apply Limits After Credibility?

A credibility-weighted indication balances a risk's observed experience with an expected or manual complement. The resulting experience modification can still imply a very large premium increase or decrease, particularly when the observed experience is unusual. A rating plan may therefore limit the final premium or limit the modification used to calculate it. The order and basis of those limits matter: a fixed dollar cap is not generally the same as a maximum experience modification.

This tutorial focuses on applying those limits after the credibility calculation. It assumes the experience rating method has already produced a credibility weight and a normalized experience measure. The central task is to keep three quantities distinct: the credibility-weighted modification, the indicated premium before limits, and the premium after limits.

Definition: Let \(P_0>0\) be the risk's premium before experience rating, \(Z\) its credibility weight with \(0\leq Z\leq1\), and \(R\) its experience ratio, defined so that \(R=1\) represents experience equal to the complement. The credibility-weighted experience modification and the indicated premium before limits are $$ M=1+Z(R-1), \qquad P_{\mathrm{ind}}=P_0M. $$ A dollar premium floor \(L\) and cap \(U\), where \(0\leq L\leq U\), produce the final premium $$ P_{\mathrm{final}}=\min\{U,\max\{L,P_{\mathrm{ind}}\}\}. $$ Alternatively, a modification floor \(m_L\) and cap \(m_U\), where \(0\leq m_L\leq m_U\), produce \(P_{\mathrm{final}}=P_0\min\{m_U,\max\{m_L,M\}\}\).

The expression \(\max\{L,P_{\mathrm{ind}}\}\) first prevents the premium from falling below its floor. Taking the minimum with \(U\) then prevents it from exceeding its cap. If the indicated premium is already between the two limits, neither changes it. If it is below the floor, the final premium is the floor; if it is above the cap, the final premium is the cap.

Calculate the Credibility Indication First

The credibility formula can be written as a weighted average:

$$ M=(1-Z)\cdot 1+Z\cdot R. $$

Thus the modification lies between the complement modification of 1 and the experience ratio \(R\), when \(Z\) is between 0 and 1. The premium limits are applied only after this blend and the multiplication by \(P_0\), unless the rating rule explicitly specifies modification limits instead.

For example, if \(R=1.8\) and \(Z=0.6\), the experience component is above the complement. The modification is \(1+0.6(1.8-1)=1.48\), not \(1.8\). A premium cap cannot be used as a substitute for calculating that credibility-weighted indication correctly.

Worked Example: Applying a Dollar Cap to a Credibility Indication

A hypothetical commercial policy has a pre-experience premium of $24,000. Its normalized experience ratio is \(R=1.8\), and its credibility weight is \(Z=0.60\). The rating rule applies a $26,000 floor and a $34,000 cap to the final premium. Calculate the final premium. Then compare it with the result of mistakenly treating 1.35 as a cap on the modification.

Method. Calculate the credibility modification, convert it to an indicated premium, and then apply the dollar limits. The common mistake to avoid is applying a modification cap when the stated limit is a dollar premium cap.

Step 1: Calculate the modification. $$ M=1+Z(R-1)=1+0.60(1.8-1)=1+0.60(0.8)=1.48. $$

Step 2: Calculate the indicated premium. $$ P_{\mathrm{ind}}=P_0M=24{,}000(1.48)=35{,}520. $$

Step 3: Apply the dollar floor and cap. The indicated premium exceeds the $34,000 cap, so $$ P_{\mathrm{final}}=\min\{34{,}000,\max\{26{,}000,35{,}520\}\}=34{,}000. $$ The floor has no effect, and the final premium is $34,000.00.

Check and trap. If 1.35 were actually a modification cap, it would produce \(24{,}000(1.35)=\$32{,}400\). That is not the stated dollar-cap result. The dollar cap is $34,000, so the correct capped premium is $34,000.00.

Dollar Limits and Modification Limits

A dollar premium limit is stated in dollars. A modification limit is a multiplier applied to the risk's own pre-experience premium. These may coincide for a particular risk, but they are not interchangeable in general. For one fixed positive \(P_0\), dollar limits \(L,U\) can be converted to modification limits \(L/P_0,U/P_0\). The proof is a direct consequence of multiplying inequalities by a positive number.

Derivation: Suppose \(P_0>0\). Multiplication by \(P_0\) preserves order, so $$ L\leq P_0M\leq U \quad\Longleftrightarrow\quad \frac{L}{P_0}\leq M\leq\frac{U}{P_0}. $$ Consequently, for this particular base premium, capping and flooring the final premium at \(L,U\) is equivalent to capping and flooring the modification at \(L/P_0,U/P_0\). If the dollar limits are fixed but \(P_0\) changes across risks, the corresponding modification limits change too. A single pair of modification limits will therefore not usually reproduce the same fixed dollar limits for every risk.

This equivalence also requires that the dollar limits be translated using the same \(P_0\) used to calculate the premium. For example, a $30,000 cap corresponds to a modification cap of 1.25 when \(P_0=\$24,000\), but to a cap of 1.00 when \(P_0=\$30,000\). A fixed modification cap, by contrast, scales with each risk's own base premium.

Worked Example: Flooring the Experience Modification

A hypothetical artisan contractor has a pre-experience premium of $42,000. Its experience ratio is \(R=0.40\), and \(Z=0.75\). The rating rule sets a modification floor of 0.70 and a modification cap of 1.30. Find the final premium, and explain why the floor is relevant.

Method. First calculate the credibility modification, then limit that modification, and finally multiply by the base premium. The common mistake is to apply the floor directly to the premium in dollars.

Step 1: Calculate the uncapped modification. $$ M=1+0.75(0.40-1)=1+0.75(-0.60)=1-0.45=0.55. $$

Step 2: Apply the modification bounds. The result 0.55 is below the 0.70 floor, so the limited modification is 0.70. The 1.30 cap has no effect.

Step 3: Calculate the final premium. $$ P_{\mathrm{final}}=42{,}000(0.70)=29{,}400. $$ The final premium is $29,400.00. Without the floor, the indication would have been \(42{,}000(0.55)=\$23,100\), confirming that the floor raises the premium by $6,300. The stated limit is 0.70 on the modification, not a $0.70 or $29,400 dollar floor.

When a Dollar Cap and a Modification Cap Diverge

A frequent design choice is whether to limit the final premium to a fixed dollar amount or limit the maximum modification. A fixed dollar cap can prevent the premium for a high-base-premium risk from rising past a particular amount. A modification cap instead limits the percentage change relative to each risk's own base premium. Neither form is automatically preferable; the rating rule must specify which one it intends.

The next example shows a fixed dollar cap in action and then contrasts it with a modification cap. It uses an explicitly stated dollar cap so that the calculation does not depend on an unstated conversion.

Worked Example: A Fixed Dollar Cap Across Different Base Premiums

A hypothetical equipment policy has \(P_0=\$24,000\), \(R=2.00\), and \(Z=0.50\). Its final premium is subject to a $27,000 floor and a $30,300 cap. Calculate the final premium. Then determine the modification that corresponds to this cap for this policy, and state what a 1.20 modification cap would produce instead.

Method. Apply the credibility calculation first, then apply the dollar limits to the resulting premium. The common mistake is to assume that a dollar cap is the same as a specified modification cap without converting it using \(P_0\).

Step 1: Calculate the credibility modification. $$ M=1+0.50(2.00-1)=1+0.50=1.50. $$

Step 2: Calculate the indicated premium. $$ P_{\mathrm{ind}}=24{,}000(1.50)=36{,}000. $$

Step 3: Apply the dollar limits. The $36,000 indication exceeds the $30,300 cap, so the final premium is $30,300.00. The $27,000 floor does not bind.

Step 4: Convert the dollar cap for this risk. $$ \frac{30{,}300}{24{,}000}=1.2625. $$ For this policy alone, a modification cap of 1.2625 is equivalent to the $30,300 cap. A modification cap of 1.20 would instead produce $$ 24{,}000(1.20)=28{,}800. $$ That result is below the dollar cap and above the dollar floor, so the final premium under a 1.20 modification cap would be $28,800.00, not $30,300.00. The two caps express different rules.

Implementation Checks and Common Traps

A reliable calculation keeps the order of operations visible. Start with the credibility inputs, calculate \(M\), calculate \(P_{\mathrm{ind}}\), and only then apply the stated limits. If the limits are on \(M\), limit \(M\) before multiplying by \(P_0\). If they are on the final premium, limit \(P_{\mathrm{ind}}\) in dollars.

1
Identify the limit basis.
Check whether the rule specifies dollar premiums or modification factors. Do not infer one from the other.
2
Calculate credibility.
Use \(M=1+Z(R-1)\), with \(R=1\) representing the complement and \(Z\) between 0 and 1.
3
Form the indicated premium.
For dollar limits, calculate \(P_{\mathrm{ind}}=P_0M\) before applying either limit.
4
Apply the specified bounds.
For either basis, a floor raises a lower result and a cap lowers a higher result. Verify which bound, if any, is binding.
5
Check the result.
The final premium must be within the stated limits, and a nonbinding limit must not change the indication.

Two exam traps deserve particular attention. First, a high experience ratio does not mean the credibility modification equals that ratio; the complement still receives weight \(1-Z\). Second, a cap on a premium is not automatically a cap on the modification. Always preserve the units: compare dollars with dollar limits and dimensionless factors with modification limits.

Key takeaway: Calculate the credibility-weighted indication before applying limits. A dollar premium cap or floor and a modification cap or floor are equivalent only when their bounds are correctly converted using the same positive base premium. Across risks with different base premiums, fixed dollar limits and fixed modification limits generally produce different results.

Check Your Understanding

For each hypothetical rating situation, determine the final premium after applying the stated limit basis. Keep the credibility calculation separate from the limiting step.

  1. A policy has \(P_0=\$18,000\), \(Z=0.40\), and \(R=1.50\). Its final premium has a $20,000 floor and a $22,000 cap. What is the final premium? (Answer: $21,600.00.)
  2. A policy has \(P_0=\$50,000\), \(Z=0.60\), and \(R=0.50\). Its modification is subject to a 0.80 floor and a 1.40 cap. What is the final premium? (Answer: $40,000.00.)
  3. A policy has \(P_0=\$32,000\), \(Z=0.50\), and \(R=2.00\). Its modification is subject to a 0.75 floor and a 1.25 cap. What is the final premium? (Answer: $40,000.00.)
  4. A policy has \(P_0=\$40,000\), \(Z=0.50\), and \(R=1.80\). Calculate the final premium first under a $50,000 dollar cap, and then under a 1.20 modification cap; there is no binding floor in either case. (Answer: $50,000.00 under the dollar cap; $48,000.00 under the modification cap.)
  5. A policy has \(P_0=\$25,000\), \(Z=0.80\), and \(R=1.10\). Its final premium has a $28,000 floor and a $35,000 cap. What is the final premium, and which limit binds? (Answer: $28,000.00; the floor binds.)
Clinical Biostats