Aggregate Deductibles with Payment Caps

How the Deductible and Cap Work Together

An aggregate deductible makes the insurer responsible only for aggregate losses above a specified threshold. A payment cap then limits how much the insurer pays, even if aggregate losses rise far beyond that threshold. The key is to translate both contract terms into a payment function before taking an expectation.

Throughout this tutorial, let $X$ be a nonnegative random variable representing aggregate loss over the contract period. Let $d\geq 0$ be the aggregate deductible, and let $u\geq 0$ be the maximum insurer payment. Assume there is no coinsurance, no discounting, and no other contract adjustment. The cap $u$ is a maximum payment, not the loss level at which the cap begins.

Definition: The insurer payment for aggregate loss $X$, with aggregate deductible $d$ and maximum payment $u$, is $Y=\min\{(X-d)_+,u\}$, where $(z)_+=\max\{z,0\}$. Equivalently, $$ Y= \begin{cases} 0, & X\leq d,\\ X-d, & d<X<d+u,\\ u, & X\geq d+u. \end{cases} $$ The cap is reached at aggregate loss $d+u$. For $u=0$, the payment is zero for every loss.

The deductible removes the first $d$ dollars of aggregate loss. Once $X$ exceeds $d$, the insurer’s payment increases dollar for dollar until it reaches $u$. Further aggregate loss is retained by the insured because the insurer has exhausted its capped payment. This is a limited aggregate excess-loss payment.

Aggregate loss	Insurer payment	Reason
$X\leq d$	0	The deductible has not been exceeded.
$d<X<d+u$	$X-d$	The payment is the amount above the deductible.
$X\geq d+u$	$u$	The maximum payment has been reached.

Expected Payment: Discrete Aggregate Loss

For a discrete aggregate-loss distribution, evaluate the payment function at each possible value of $X$, then weight each payment by its probability. The payment function is nonlinear, so in general it is not correct to apply the deductible and cap to the expected loss alone.

Worked Example: A Cap Applied to a Discrete Loss Table

A hypothetical group contract has aggregate loss $X$ equal to $\$0$, $\$1{,}000$, $\$3{,}000$, or $\$6{,}000$, with respective probabilities $0.20$, $0.30$, $0.30$, and $0.20$. The aggregate deductible is $d=\$1{,}500$, and the maximum insurer payment is $u=\$2{,}500$. Find the expected insurer payment.

Method. First calculate $d+u=\$1{,}500+\$2{,}500=\$4{,}000$. Losses at or below $\$1{,}500$ produce no payment; losses from above $\$1{,}500$ up to $\$4{,}000$ produce the excess over the deductible; losses at or above $\$4{,}000$ produce the capped payment of $\$2{,}500$.

At $X=\$0$, the payment is $\min\{(0-1{,}500)_+,2{,}500\}=\$0$. At $X=\$1{,}000$, it is $\min\{0,2{,}500\}=\$0$. At $X=\$3{,}000$, it is $\min\{3{,}000-1{,}500,2{,}500\}=\$1{,}500$. At $X=\$6{,}000$, it is $\min\{6{,}000-1{,}500,2{,}500\}=\$2{,}500$.

\begin{aligned} E[Y] &=(0.20)(0)+(0.30)(0)+(0.30)(1{,}500)+(0.20)(2{,}500)\\ &=0+0+450+500\\ &=\$950. \end{aligned}

The probabilities sum to $0.20+0.30+0.30+0.20=1.00$, and the expected payment, $\$950$, is between zero and the cap of $\$2{,}500$. Common mistake avoided: using $\$2{,}500$ as the loss threshold for the cap. It is the payment cap; the insurer reaches it when $X=d+u=\$4{,}000$.

A Tail-Integral Formula and Why It Holds

When the aggregate-loss distribution is continuous or described by a survival function, an integral can be more convenient than enumerating loss values. Let $S_X(t)=P(X>t)$ be the survival function of $X$. For nonnegative $X$, the expected capped payment is the area under the survival curve between the deductible and the loss level where the cap is reached.

Theorem: For $d\geq 0$, $u\geq 0$, and nonnegative aggregate loss $X$, $$ E[\min\{(X-d)_+,u\}] =\int_d^{d+u} S_X(t)\,dt =E[(X-d)_+]-E[(X-d-u)_+]. $$ The identity holds whether $X$ is discrete, continuous, or mixed.

To see the integral identity, write the payment as the length of the portion of the interval $[d,d+u]$ that lies below $X$. For each fixed value of $X$, this length is zero if $X\leq d$, is $X-d$ if $d<X<d+u$, and is $u$ if $X\geq d+u$. Thus it equals $\min\{(X-d)_+,u\}$. Taking expectations and integrating the indicator of $X>t$ over the interval gives the tail integral. Whether an endpoint is included does not change an integral over $t$; in a discrete model the formula still correctly accounts for probability atoms.

\begin{aligned} \min\{(X-d)_+,u\} &=\int_d^{d+u}\mathbf{1}_{\{X>t\}}\,dt,\\ E[\min\{(X-d)_+,u\}] &=\int_d^{d+u}E[\mathbf{1}_{\{X>t\}}]\,dt\\ &=\int_d^{d+u}P(X>t)\,dt =\int_d^{d+u}S_X(t)\,dt. \end{aligned}

For the difference formula, use the limited-payment identity $\min\{z,u\}=z-(z-u)_+$ for $z\geq0$, with $z=(X-d)_+$. Then $(z-u)_+=(X-d-u)_+$, giving the stated difference of two stop-loss expectations. This form is useful when those two expectations are already available.

Worked Example: Uniform Aggregate Loss

A hypothetical contract’s aggregate loss is uniformly distributed on $[0,\$10{,}000]$. Thus $f_X(x)=1/10{,}000$ on that interval and zero elsewhere. The deductible is $d=\$2{,}000$, and the maximum insurer payment is $u=\$3{,}000$. Find the expected payment.

Method. The payment increases from zero to $\$3{,}000$ as $X$ rises from $\$2{,}000$ to $d+u=\$5{,}000$. Above $\$5{,}000$, it remains at the cap. Equivalently, for $0\leq t\leq \$10{,}000$, $S_X(t)=(10{,}000-t)/10{,}000$. The tail-integral limits are $d=\$2{,}000$ and $d+u=\$5{,}000$.

\begin{aligned} E[Y] &=\int_{2{,}000}^{5{,}000}\frac{10{,}000-t}{10{,}000}\,dt\\ &=\frac{1}{10{,}000} \left[10{,}000t-\frac{t^2}{2}\right]_{2{,}000}^{5{,}000}\\ &=\frac{1}{10{,}000} \left[(50{,}000{,}000-12{,}500{,}000) -(20{,}000{,}000-2{,}000{,}000)\right]\\ &=\frac{19{,}500{,}000}{10{,}000} =\$1{,}950. \end{aligned}

As a check using the payment curve, the rising section is a triangle with area $\tfrac12(3{,}000)(3{,}000)=4{,}500{,}000$ dollar-squared units. The flat section runs from $\$5{,}000$ to $\$10{,}000$, with area $(5{,}000)(3{,}000)=15{,}000{,}000$. Dividing their sum, $19{,}500{,}000$, by the uniform interval length $\$10{,}000$ again gives $\$1{,}950$. Common mistake avoided: treating the cap as an aggregate-loss limit of $\$3{,}000$. The payment reaches its cap at a loss of $\$5{,}000$, and remains there for larger losses.

Applying the Contract to a Compound Aggregate Loss

An aggregate loss may be built from a random number of individual losses. The contract payment is still determined by the realized total $X$, not by applying a separate deductible or cap to each individual loss. A useful sequence is to obtain the possible aggregate totals, map each total through the payment function, and then take the expectation.

Worked Example: Binomial Claim Count with Fixed Claim Amounts

In a hypothetical one-year portfolio, the claim count satisfies $N\sim\operatorname{Binomial}(5,0.40)$, and each claim is exactly $\$1{,}000$. Assume claim count is the only source of randomness, so $X=\$1{,}000N$. The aggregate deductible is $d=\$1{,}000$, and the maximum insurer payment is $u=\$2{,}000$. Find the expected payment.

Method. The cap is reached at $d+u=\$3{,}000$. If $N=0$ or $1$, then $X\leq\$1{,}000$ and the payment is zero. If $N=2$, then $X=\$2{,}000$, so the payment is $\$1{,}000$. If $N\geq3$, the excess over the deductible is at least $\$2{,}000$, so the payment is capped at $\$2{,}000$.

The probability of exactly two claims is $P(N=2)=\binom52(0.40)^2(0.60)^3=10(0.16)(0.216)=0.3456$. The probability of three or more claims is $P(N\geq3)=P(N=3)+P(N=4)+P(N=5)$, where $P(N=3)=10(0.40)^3(0.60)^2=0.2304$, $P(N=4)=5(0.40)^4(0.60)=0.0768$, and $P(N=5)=(0.40)^5=0.01024$. Therefore $P(N\geq3)=0.31744$.

\begin{aligned} E[Y] &=0\cdot P(N\leq1) +1{,}000P(N=2)+2{,}000P(N\geq3)\\ &=(1{,}000)(0.3456)+(2{,}000)(0.31744)\\ &=345.60+634.88\\ &=\$980.48. \end{aligned}

The probabilities used for positive payments total $0.3456+0.31744=0.66304$, leaving $0.33696$ for $N\leq1$; together they sum to $1$. The expected payment is below the maximum payment, as required. Common mistake avoided: counting the $N=1$ outcome as a $\$1{,}000$ payment. One claim exactly exhausts the deductible, so it produces no insurer payment.

Common Errors and Practical Checks

An accurate expected payment begins with a correctly interpreted contract. Several checks help catch errors before they affect the final calculation.

Confusing the payment cap with the loss threshold. If the deductible is $d$ and the payment cap is $u$, the cap is reached at $X=d+u$, not at $X=u$.
Applying the deductible more than once. An aggregate deductible is applied to the total $X$. It is not automatically a per-claim deductible.
Using an uncapped excess payment. The expression $(X-d)_+$ ignores the payment cap. The capped payment is $\min\{(X-d)_+,u\}$.
Using $E[X]$ in place of the distribution of $X$. In general, $E[\min\{(X-d)_+,u\}]$ is not $\min\{(E[X]-d)_+,u\}$. The contract payment is nonlinear, so the loss distribution matters.
Misclassifying boundary outcomes. At $X=d$, payment is zero; at $X=d+u$, payment is exactly $u$. Check these points explicitly when the aggregate distribution has atoms.

A quick bound is $0\leq E[Y]\leq u$. Also, since $Y\leq(X-d)_+$, its expectation cannot exceed the uncapped stop-loss expectation at $d$. These are useful checks, but they do not replace calculating the payment under the full contract.

Key takeaway: For each aggregate loss, subtract the deductible, floor the result at zero, and then limit the payment to $u$. The expected capped payment is the weighted average of those payments, or equivalently the integral of $S_X(t)$ from $d$ to $d+u$. This treatment of aggregate deductibles and caps is aligned with the SOA Exam FAM (Fundamentals of Actuarial Mathematics) syllabus.

Check Your Understanding

For each problem, determine the insurer payment after applying both the aggregate deductible and the payment cap.

A hypothetical aggregate loss is $\$0$, $\$2{,}000$, or $\$4{,}000$, with respective probabilities $0.10$, $0.50$, and $0.40$. The deductible is $\$1{,}000$, and the maximum payment is $\$2{,}000$. Find the expected payment. (Answer: $\$1{,}300$)
A hypothetical aggregate loss is uniform on $[0,\$12{,}000]$. The deductible is $\$3{,}000$, and the maximum payment is $\$3{,}000$. Find the expected payment. (Answer: $\$1{,}875$)
A discrete aggregate loss takes values $\$0,\$1{,}000,\$2{,}500,\$4{,}000,\$7{,}000$, each with probability $0.20$. The deductible is $\$1{,}500$, and the maximum payment is $\$2{,}500$. Find the expected payment. (Answer: $\$1{,}200$)
A hypothetical claim count satisfies $N\sim\operatorname{Binomial}(4,0.50)$, with each claim exactly $\$750$, so $X=\$750N$. The deductible is $\$750$, and the maximum payment is $\$1{,}500$. Find the expected payment. (Answer: $\$750$)
A hypothetical aggregate loss has values $\$0,\$2{,}000,\$5{,}000,\$8{,}000$, each with probability $0.25$. The deductible is $\$2{,}000$, and the maximum payment is $\$3{,}000$. Find the expected payment, paying particular attention to the outcome exactly at the deductible. (Answer: $\$1{,}500$)

Aggregate loss	Insurer payment	Reason
\(X\leq d\)	0	The deductible has not been exceeded.
\(d<X<d+u\)	\(X-d\)	The payment is the amount above the deductible.
\(X\geq d+u\)	\(u\)	The maximum payment has been reached.

What You'll Learn