Assess Incentives Created By Aggregate Caps

Why an Aggregate Cap Changes Marginal Premium

In a retrospective plan, the premium may change with the insured’s aggregate losses. If the plan also caps the loss-sensitive component, the premium effect of an additional loss depends on how much of that cap has already been used. An additional loss can affect the premium in full, affect it only up to the remaining cap, or have no effect on that component.

The calculation must follow the contract’s order of operations. In this tutorial, the aggregate cap applies to the converted aggregate loss component: first multiply aggregate eligible losses by the loss conversion factor, then apply the cap. This is distinct from a contract that caps losses before applying the conversion factor.

Definition: Let $L$ be aggregate eligible losses under the plan, $k$ the loss conversion factor (LCF), $C$ the maximum amount of the converted loss component, and $B$ the fixed basic-premium component. Under the simplified plan considered here, with no minimum or maximum premium also binding, $$ P(L)=B+\min(kL,C). $$ The aggregate loss threshold at which the cap first binds is $L^*=C/k$, for $k>0$. The loss-sensitive component is $kL$ below this threshold and $C$ at or above it.

Here, $B$ and $C$ are held fixed as losses change. The eligible loss amount $L$ is the amount specified by the plan; any applicable claim-level limits or other adjustments must be reflected in $L$ before using this formula. The examples use hypothetical plans and nominal loss amounts at the same adjustment basis. No interest, mortality, or probability distribution is assumed: these are deterministic premium calculations.

Marginal Impact of Additional Losses

Suppose aggregate eligible losses increase from $L$ to $L+\Delta L$, where $\Delta L\geq 0$. The premium change is the difference between the capped loss components, not automatically the conversion factor times the entire additional loss:

\Delta P =\min\bigl(k(L+\Delta L),C\bigr)-\min(kL,C).

If losses are below the cap both before and after the increase, each additional dollar of eligible loss changes the premium by $k$ dollars. If the increase crosses the threshold, only the part of the increase up to the threshold affects the premium. If losses are already at or above the threshold, further losses do not change this capped component.

\Delta P= \begin{cases} k\Delta L, & L+\Delta L\leq L^*,\\ C-kL, & L<L^*<L+\Delta L,\\ 0, & L\geq L^*. \end{cases}

At exactly $L^*$, the premium formula has a change in slope: immediately below the threshold, the marginal effect is $k$; immediately above it, the effect is zero. For a finite loss increase that ends exactly at the threshold, the full increase still affects premium, since $k(L+\Delta L)=C$.

Why the Formula Works

The cap limits the total converted loss component over the portfolio, not the effect of each claim considered separately. The converted amount before capping is $kL$. If that amount is less than $C$, the plan charges that amount; if it exceeds $C$, the plan charges only $C$. That is exactly the operation $\min(kL,C)$.

For the marginal calculation, subtract the loss component at the original aggregate from the loss component after the increase. If the new converted amount remains below the cap, the difference is $k(L+\Delta L)-kL=k\Delta L$. If it crosses the cap, the new component is $C$, so the difference is $C-kL$. If the original amount is already at the cap, both the original and new components equal $C$, so the difference is zero. This also explains why multiplying every additional loss by $k$ overstates the premium change when the increase crosses or lies beyond the cap.

Worked Calculations

Worked Example: An Increase That Crosses the Cap

A hypothetical retrospective plan has basic premium $B=\$240{,}000$, LCF $k=1.25$, and aggregate converted-loss cap $C=\$900{,}000$. Aggregate eligible losses are currently $\$680{,}000$ and then increase by $\$80{,}000$. Calculate the premium before and after the increase and the premium impact. All amounts are nominal dollars at the same adjustment basis; no interest, mortality, or loss distribution is assumed.

Method. Calculate the threshold, then apply the cap to converted aggregate losses at each loss level.

The loss threshold is:

L^*=\frac{C}{k} =\frac{\$900{,}000}{1.25} =\$720{,}000.

Before the increase, converted losses are $1.25\times \$680{,}000=\$850{,}000$, below the cap. Thus:

P(\$680{,}000) =\$240{,}000+\min(\$850{,}000,\$900{,}000) =\$1{,}090{,}000.

After the increase, eligible losses are $\$680{,}000+\$80{,}000=\$760{,}000$. Converted losses are $1.25\times \$760{,}000=\$950{,}000$, so the capped component is $\$900{,}000$:

P(\$760{,}000) =\$240{,}000+\min(\$950{,}000,\$900{,}000) =\$1{,}140{,}000.

The premium increase is $\$1{,}140{,}000-\$1{,}090{,}000=\$50{,}000$. Only $\$720{,}000-\$680{,}000=\$40{,}000$ of the additional eligible losses fit below the threshold, and $1.25\times \$40{,}000=\$50{,}000$. The two calculations agree.

Common mistake avoided: Multiplying the full $\$80{,}000$ increase by 1.25 gives a wrong impact of $\$100{,}000$. That ignores the cap. The calculation is checked by comparing the two capped premiums: their difference is $\$50{,}000$, not $\$100{,}000$.

Worked Example: A Loss Increase Partly Uses the Remaining Cap

A hypothetical plan has $B=\$175{,}000$, $k=0.80$, and $C=\$640{,}000$. Aggregate eligible losses of $\$750{,}000$ increase by $\$100{,}000$. Find the premium change and the amount of additional eligible losses that affects it. Treat all amounts as nominal dollars at the same adjustment basis; no interest, mortality, or probability distribution is assumed.

The threshold is:

L^*=\frac{\$640{,}000}{0.80}=\$800{,}000.

Before the increase, converted losses are $0.80\times \$750{,}000=\$600{,}000$, below the cap. The premium is:

P(\$750{,}000) =\$175{,}000+\min(\$600{,}000,\$640{,}000) =\$775{,}000.

After the increase, eligible losses are $\$850{,}000$, and converted losses are $0.80\times \$850{,}000=\$680{,}000$. The component is capped at $\$640{,}000$, so:

P(\$850{,}000) =\$175{,}000+\min(\$680{,}000,\$640{,}000) =\$815{,}000.

The increase in premium is $\$815{,}000-\$775{,}000=\$40{,}000$. The plan has room for $\$800{,}000-\$750{,}000=\$50{,}000$ of eligible losses before reaching the threshold. Those losses add $0.80\times \$50{,}000=\$40{,}000$ to premium; the remaining $\$50{,}000$ of the loss increase has no further effect on the capped component.

Common mistake avoided: Applying the cap to the $\$100{,}000$ increase alone would not use the plan’s cumulative loss position. The result is checked by evaluating the premium at both aggregate loss amounts: the difference is $\$40{,}000$.

Worked Example: Comparing Marginal Impacts Under Two Plans

A hypothetical insured has $B=\$200{,}000$ and aggregate eligible losses of $\$790{,}000$. Losses then increase by $\$40{,}000$. Compare the premium impact under Plan A, with $k=1.20$ and $C=\$960{,}000$, and Plan B, with $k=1.00$ and $C=\$900{,}000$. All amounts are nominal dollars at the same adjustment basis; no interest, mortality, or loss distribution is assumed.

For Plan A, the threshold is $\$960{,}000/1.20=\$800{,}000$. Before the increase, the converted amount is $1.20\times \$790{,}000=\$948{,}000$, so premium is $\$200{,}000+\$948{,}000=\$1{,}148{,}000$. After the increase, losses are $\$830{,}000$, the converted amount is $1.20\times \$830{,}000=\$996{,}000$, and the capped component is $\$960{,}000$. Premium is $\$200{,}000+\$960{,}000=\$1{,}160{,}000$. Plan A’s premium impact is $\$12{,}000$.

For Plan B, the threshold is $\$900{,}000/1.00=\$900{,}000$. Before the increase, the converted amount is $\$790{,}000$, giving premium of $\$200{,}000+\$790{,}000=\$990{,}000$. After the increase, the converted amount is $\$830{,}000$, still below the cap, giving premium of $\$200{,}000+\$830{,}000=\$1{,}030{,}000$. Plan B’s premium impact is $\$40{,}000$.

Although Plan A has the larger LCF, its cap is much closer to binding at the starting loss level, so its premium impact is smaller. The check is the difference between each plan’s computed premiums: $\$12{,}000$ for Plan A and $\$40{,}000$ for Plan B.

Common mistake avoided: Comparing LCFs alone does not determine the marginal impact when caps differ. The result depends on both the conversion factor and the remaining room under the cap.

Interpreting the Incentive

The cap changes the plan’s marginal premium incentive as cumulative losses rise. Before the threshold, an additional eligible loss increases the loss component at the LCF. Once the threshold is reached, the capped component no longer increases. For a loss increase that crosses the threshold, the increase in premium reflects only the remaining room under the cap.

This is a statement about the premium formula, not a claim that the cap changes the underlying frequency or severity of losses. It also does not determine the total premium impact if some other plan feature changes at the same time. If a contract includes a minimum premium, a maximum premium, or a different order of operations, those features must be included in the calculation. In particular, if the contract caps eligible losses before conversion, the loss threshold and marginal impact must be calculated from that contract’s formula rather than from $B+\min(kL,C)$.

Use cumulative eligible losses to determine whether the cap is already binding.
Confirm whether the cap applies before or after the LCF.
For a crossing increase, calculate the remaining room under the cap rather than charging the full increase at the LCF.
Check the answer by calculating the premium at both loss levels and subtracting.

Key takeaway: Under a cap on converted aggregate losses, the premium effect of additional losses is $k$ times the eligible loss increase only while the entire increase fits below the threshold. A crossing increase uses only the remaining cap room; losses added after the cap binds have zero impact on this component.

Check Your Understanding

For each hypothetical plan, assume the stated cap applies to converted aggregate losses, the basic premium is fixed, and no other premium limit binds. Amounts are nominal dollars at the same adjustment basis; no interest, mortality, or probability distribution is assumed.

A plan has $k=1.10$, $C=\$550{,}000$, and current eligible losses of $\$480{,}000$. Losses increase by $\$40{,}000$. What is the premium impact? (Answer: $\$22{,}000$)
A plan has $k=0.90$, $C=\$720{,}000$, and current eligible losses of $\$760{,}000$. Losses increase by $\$50{,}000$. What is the premium impact? (Answer: $\$36{,}000$)
A plan has $k=1.50$, $C=\$1{,}200{,}000$, and current eligible losses of $\$790{,}000$. Losses increase by $\$30{,}000$. What is the premium impact? (Answer: $\$15{,}000$)
A plan has $k=1.20$, $C=\$1{,}000{,}000$, and current eligible losses of $\$900{,}000$. Losses increase by $\$10{,}000$. What is the impact on the capped loss component? (Answer: $\$0$)
For the same starting eligible losses of $\$600{,}000$ and increase of $\$100{,}000$, compare Plan A ($k=1.30,\ C=\$900{,}000$) with Plan B ($k=1.10,\ C=\$800{,}000$). What is the premium impact under each plan? (Answer: Plan A, $\$120{,}000$; Plan B, $\$110{,}000$)

What You'll Learn