Expected Insurer Loss by Deductible Layer

Why Expected Loss by Layer Matters

A deductible allocates each covered claim between the insured and the insurer. For pricing or comparing coverage structures, it is often useful to calculate the expected amount allocated to each party from a claim severity distribution. The key is to calculate the expected payment for the particular layer—not to mistake the insurer’s payment for the total expected claim.

This tutorial concerns one covered claim at a time. It does not incorporate claim frequency, annual aggregate deductibles, expenses unless they are included in the stated severity, or discounting. All scenarios are hypothetical. Unless a problem says otherwise, assume severity is measured on the same dollar basis as the deductible and limit, claims are nonnegative, and there is no interest or mortality assumption because neither applies to this single-claim calculation.

Definition: Let $X$ be the ground-up severity of one covered claim, before applying the deductible. With a per-claim deductible $d$, the insured’s retained payment is $\min(X,d)$, and the insurer’s payment with no upper limit is $(X-d)_+=\max(X-d,0)$. Thus $X=\min(X,d)+(X-d)_+$ for every claim.

The insurer pays nothing when $X\le d$. When $X>d$, the insurer pays the amount by which the claim exceeds $d$. Consequently, the probability of an insurer payment is $P(X>d)$, but that probability alone does not give the expected payment: claim amounts above the deductible also matter.

Expected Payment in the Deductible Layers

Write $S_X(x)=P(X>x)$ for the severity survival function. For a nonnegative claim, the expected retained amount below the deductible and the expected unlimited insurer payment above it can be calculated as layer integrals:

E[\min(X,d)]=\int_0^d S_X(t)\,dt, \qquad E[(X-d)_+]=\int_d^\infty S_X(t)\,dt.

These formulas apply when the corresponding expectations exist. They are not restricted to continuous distributions; atoms, including a probability mass at the deductible, are handled by the survival-function integrals. For a severity with a density $f_X$, the insurer expectation can also be calculated directly as $\int_d^\infty (x-d)f_X(x)\,dx$. The survival form is often simpler, particularly when the survival function has a convenient expression.

Layer identity: Since $X=\min(X,d)+(X-d)_+$, a finite expected ground-up loss satisfies $E[X]=E[\min(X,d)]+E[(X-d)_+]$. This equality is a useful reconciliation check. It does not mean that either layer expectation equals $E[X]$ by itself.

Why the Layer Integral Works

For any fixed nonnegative amount $y$, the length of the interval from zero to $y$ is $y$. Equivalently, $y=\int_0^\infty \mathbf{1}(y>t)\,dt$, where the indicator is one when its condition holds and zero otherwise. Applying this identity to the amount paid above the deductible gives:

(X-d)_+=\int_d^\infty \mathbf{1}(X>t)\,dt.

Taking expectations and interchanging the expectation and integral for a nonnegative integrand yields:

E[(X-d)_+] =\int_d^\infty E[\mathbf{1}(X>t)]\,dt =\int_d^\infty P(X>t)\,dt =\int_d^\infty S_X(t)\,dt.

The same argument over the interval from zero to $d$ gives $\min(X,d)=\int_0^d \mathbf{1}(X>t)\,dt$, and hence the retained-layer formula. This derivation explains why the insurer’s expected payment depends on the entire tail of the severity distribution above $d$, not only on the chance that the deductible is exceeded.

Worked Calculations

Worked Example: Uniform Severity Above a Deductible

A hypothetical insurer covers one claim with ground-up severity $X\sim U(0,20{,}000)$. The per-claim deductible is $d=\$8{,}000$, with no insurer payment limit. Find the probability that the insurer pays, the expected insurer payment, and the expected insured retention. Assume no discounting (no interest adjustment); mortality is not applicable. The severity is uniform on the stated interval.

Method. For this uniform distribution, $S_X(t)=1-t/20{,}000$ for $0\le t\le20{,}000$. The insurer’s probability of paying is:

P(X>8{,}000)=1-\frac{8{,}000}{20{,}000}=0.6.

Its expected payment is the survival area above the deductible. The survival function is zero beyond $\$20{,}000$, so:

\begin{aligned} E[(X-8{,}000)_+] &=\int_{8{,}000}^{20{,}000}\left(1-\frac{t}{20{,}000}\right)\,dt\\ &=\left[t-\frac{t^2}{40{,}000}\right]_{8{,}000}^{20{,}000}\\ &= \left(20{,}000-10{,}000\right) -\left(8{,}000-1{,}600\right)\\ &=10{,}000-6{,}400\\ &=\$3{,}600. \end{aligned}

The expected ground-up severity of a uniform distribution on zero to $\$20{,}000$ is $(0+20{,}000)/2=\$10{,}000$. Therefore the expected insured retention is $\$10{,}000-\$3{,}600=\$6{,}400$. As a direct check, $E[\min(X,8{,}000)]=\int_0^{8{,}000}(1-t/20{,}000)\,dt=\$6{,}400$. The two expected layers sum to $\$10{,}000$, as required. The common mistake avoided here is evaluating the lower endpoint incorrectly: $8{,}000-8{,}000^2/40{,}000=8{,}000-1{,}600=6{,}400$, not $8{,}400$.

Worked Example: Exponential Severity and a Deductible

For a hypothetical single claim, let $X$ have an exponential distribution with mean $\theta=\$12{,}000$, so $S_X(x)=e^{-x/12{,}000}$ for $x\ge0$. The deductible is $d=\$5{,}000$, and the insurer layer is unlimited. Find the probability of an insurer payment and its expected amount, then calculate expected insured retention. There is no discounting or mortality assumption.

Method. Integrating the exponential survival function gives the expected payment:

\begin{aligned} E[(X-d)_+] &=\int_d^\infty e^{-t/\theta}\,dt\\ &=\theta e^{-d/\theta}\\ &=12{,}000e^{-5{,}000/12{,}000}\\ &=12{,}000e^{-5/12}\\ &\approx 12{,}000(0.6592406302)\\ &=\$7{,}910.89. \end{aligned}

The probability of payment is $P(X>5{,}000)=e^{-5/12}\approx0.6592$, to four decimal places. The expected ground-up severity is the exponential mean, $\$12{,}000$, so expected insured retention is $\$12{,}000-\$7{,}910.89=\$4{,}089.11$. The check is $\$7{,}910.89+\$4{,}089.11=\$12{,}000.00$. The probability $0.6592$ is not itself the expected payment; multiplying it by the mean $\$12{,}000$ would incorrectly treat the conditional excess amount as $\$12{,}000$.

Worked Example: A Capped Insurer Layer Under Pareto Severity

A hypothetical claim has Lomax (Pareto Type II) severity with survival function $S_X(x)=(1+x/\theta)^{-\alpha}$, where $\theta=\$10{,}000$ and $\alpha=2$, for $x\ge0$. The policy has a $\$10{,}000$ deductible and an insurer payment cap of $\$20{,}000$ per claim. Calculate the expected insurer payment and expected insured retention. There is no discounting or mortality assumption.

Method. The insurer pays $\min((X-d)_+,20{,}000)$. The cap means the insurer’s layer ends at ground-up severity $d+20{,}000=\$30{,}000$. Thus its expectation is the survival area from $d$ to $\$30{,}000$:

\begin{aligned} E[\min((X-10{,}000)_+,20{,}000)] &=\int_{10{,}000}^{30{,}000} \left(1+\frac{t}{10{,}000}\right)^{-2}\,dt\\ &=10{,}000\left[ \frac{1}{1+10{,}000/10{,}000} -\frac{1}{1+30{,}000/10{,}000} \right]\\ &=10{,}000\left(\frac12-\frac14\right)\\ &=\$2{,}500. \end{aligned}

The distribution’s mean is $\theta/(\alpha-1)=\$10{,}000$, so expected insured retention, including amounts above the insurer cap, is $\$10{,}000-\$2{,}500=\$7{,}500$. To check the layer boundary, the expected insurer payment in the uncapped layer above the deductible would be $\int_{10{,}000}^{\infty}S_X(t)\,dt=\$5{,}000$. The portion above ground-up severity $\$30{,}000$ is another $\int_{30{,}000}^{\infty}S_X(t)\,dt=\$2{,}500$, so the cap reduces the insurer expectation from $\$5{,}000$ to $\$2{,}500$. The two actual expected payments, $\$2{,}500$ and $\$7{,}500$, reconcile to the $\$10{,}000$ mean.

Limits, Interpretation, and Exam Traps

A per-claim deductible and an insurer payment cap define different ends of the insurer’s layer. If the insurer pays above deductible $d$, with a maximum payment $c$, the payment is $\min((X-d)_+,c)$. Its expected value is:

E[\min((X-d)_+,c)]=\int_d^{d+c} S_X(t)\,dt.

The upper integration endpoint is $d+c$, not $c$, because $c$ is the maximum insurer payment in excess of the deductible. If a contract instead specifies a ground-up maximum severity $u$, with $u>d$, the equivalent cap is $c=u-d$, and the insurer layer expectation is $\int_d^u S_X(t)\,dt$. Contract wording determines which interpretation applies.

A useful calculation sequence is to identify the ground-up severity distribution, mark the layer boundaries in ground-up dollars, and then integrate the survival function over the insurer’s actual payment layer. Separately calculate the expected total severity or retained layer if needed, and reconcile the parts. This makes the result easier to interpret and helps catch a deductible or limit applied at the wrong point.

Do not use exceedance probability as expected loss. $P(X>d)$ counts how often the insurer pays; the expected payment also reflects how far claims exceed $d$.
Do not integrate the claim amount in the insurer layer without subtracting the deductible. For a density calculation, the unlimited insurer payment is $\int_d^\infty(x-d)f_X(x)\,dx$, not $\int_d^\infty xf_X(x)\,dx$.
Do not omit the capped tail incorrectly. In a direct payment calculation, claims above $d+c$ still produce payment $c$; they do not produce zero. The survival-layer integral to $d+c$ includes that payment correctly.
Keep the layer and unit consistent. A cap on insurer payment is an amount above the deductible; a ground-up limit is measured from zero. Do not treat these as interchangeable without converting $u=d+c$.

Key takeaway: The expected payment in a layer is the area under the severity survival curve across that layer. Define the ground-up boundaries first, use the correct survival function, and verify that the expected allocated amounts reconcile to the expected ground-up loss when the layers cover the full claim.

Check Your Understanding

For each hypothetical single-claim scenario, identify the insurer layer and calculate the requested expected payment. Round dollar answers to the nearest cent where needed; give probabilities to four decimals.

Let $X\sim U(0,30{,}000)$ with a $\$9{,}000$ deductible and no insurer cap. Calculate the probability of an insurer payment and the expected insurer payment. (Answer: probability $0.7000$; expected insurer payment $\$7{,}350.00$)
Let $X$ be exponential with mean $\$8{,}000$. For a $\$4{,}000$ deductible and no cap, calculate expected insurer payment and expected insured retention. (Answer: insurer $\$4{,}852.25$; insured $\$3{,}147.75$)
Let $X$ have Lomax survival $S_X(x)=(1+x/5{,}000)^{-3}$, $x\ge0$. With a $\$5{,}000$ deductible and no cap, calculate the expected insurer payment. (Answer: $\$625.00$)
A severity distribution has $P(X=0)=0.5$, $P(X=4{,}000)=0.3$, and $P(X=12{,}000)=0.2$. With a $\$5{,}000$ deductible and no cap, calculate expected insurer payment and expected insured retention. (Answer: insurer $\$1{,}400.00$; insured $\$2{,}200.00$)
Let $X\sim U(0,40{,}000)$, with a $\$10{,}000$ deductible and a $\$25{,}000$ maximum insurer payment. Calculate the expected insurer payment, including claims that exceed the insurer’s layer. (Answer: $\$10{,}937.50$)

What You'll Learn