Losses Crossing Multiple Excess of Loss Layers

Allocating One Loss Across Consecutive Layers

A large loss may extend through several excess of loss layers. To allocate it correctly, compare the same subject loss with each layer’s attachment point and upper boundary. A lower layer can be exhausted while a higher layer responds only partially; a loss above the program’s top can leave the cedent with an amount retained there as well.

The key discipline is to calculate each layer’s recovery from the original subject loss, not from a balance successively reduced by earlier recoveries. The layers cover different bands of loss, so their recoveries add without paying the same band twice.

Definition: A per-loss excess of loss layer written $L$ excess of $A$ has attachment point $A$ and limit $L$. For a nonnegative subject loss $X$, its recovery is $R(A,L;X)=\min\{\max(X-A,0),L\}$. The layer’s upper boundary is $A+L$. It is unentered when $X\leq A$, partially used when $A<X<A+L$, and exhausted when $X\geq A+L$.

For example, a layer of $400,000 excess of $600,000 attaches at $600,000 and covers the next $400,000 of loss, up to a boundary of $1,000,000. A $750,000 subject loss produces a $150,000 recovery from that layer. A loss of $1,000,000 or more exhausts it; the layer’s recovery cannot exceed its $400,000 limit.

In this tutorial, assume each stated layer applies to the same single subject loss, in the same currency and on the same coverage basis. Ignore exclusions, expenses, reinstatements, and other reinsurance adjustments. The layers are consecutive: each higher layer attaches exactly at the lower layer’s upper boundary. These assumptions make it possible to allocate the loss by its amount alone.

Work From Attachment Points and Boundaries

For each layer, record its attachment point $A_j$, limit $L_j$, and upper boundary $U_j=A_j+L_j$. In a consecutive program, the next layer attaches at $U_j$. Then compare $X$ with the two boundaries. The recovery is zero below attachment, the amount by which the loss exceeds attachment while it remains within the layer, or the full limit once the loss reaches or exceeds the upper boundary.

R_j=\min\{\max(X-A_j,0),L_j\}

The positive-part operation $\max(X-A_j,0)$ removes any amount of loss below the layer’s attachment. Taking the minimum with $L_j$ caps the recovery at the layer limit. Thus, the formula handles all three cases without applying the limit incorrectly.

Worked Example: A Loss Exhausts Two Layers and Enters a Third

A hypothetical insurer has three consecutive layers: $500,000 excess of $500,000; $1,000,000 excess of $1,000,000; and $1,000,000 excess of $2,000,000. A single subject loss is $2,400,000. Find each recovery, identify which layers are exhausted, and calculate the amount retained by the cedent. Assume there are no other recoveries or adjustments.

Method. Find each layer’s interval first, then apply the recovery formula to the original $2,400,000 loss. All amounts below are in dollars.

First layer. Its attachment is $500,000 and its upper boundary is $500,000 + $500,000 = $1,000,000. Therefore, $R_1=\min\{\max(2{,}400{,}000-500{,}000,0),500{,}000\}=\min(1{,}900{,}000,500{,}000)=500{,}000.$ The layer is exhausted because the loss exceeds its $1,000,000 upper boundary.

Second layer. It attaches at $1,000,000 and ends at $1,000,000 + $1,000,000 = $2,000,000. Its recovery is $R_2=\min\{\max(2{,}400{,}000-1{,}000{,}000,0),1{,}000{,}000\}=\min(1{,}400{,}000,1{,}000{,}000)=1{,}000{,}000.$ This layer is also exhausted.

Third layer. It attaches at $2,000,000 and ends at $3,000,000. Its recovery is $R_3=\min\{\max(2{,}400{,}000-2{,}000{,}000,0),1{,}000{,}000\}=\min(400{,}000,1{,}000{,}000)=400{,}000.$ This layer is partially used, not exhausted.

Total recovery and check. The recoveries total $500,000 + $1,000,000 + $400,000 = $1,900,000. The cedent retains $2,400,000 − $1,900,000 = $500,000, which is the loss below the program’s $500,000 attachment. The check is $1,900,000 recovered + $500,000 retained = $2,400,000.

Common mistake avoided. Do not subtract the first recovery from the loss and then apply the second layer’s attachment to that remainder. Each layer responds to the part of the original loss in its own interval.

Why Consecutive Layer Recoveries Add

A layer covers only the band from its attachment to its upper boundary. If consecutive layers cover adjacent bands, their covered portions together form one continuous interval. This gives a useful check on a layer-by-layer calculation.

Write the boundaries.
Let the consecutive layers cover $[A_0,A_1]$, $[A_1,A_2]$, through $[A_{n-1},A_n]$. Each layer’s limit is the width of its interval.

Identify the loss portion in each interval.
For a nonnegative loss $X$, the amount in layer $j$ is the part of $X$ above $A_{j-1}$, capped at $A_j-A_{j-1}$.

Add the adjacent portions.
The intervals do not overlap, so their covered portions join from $A_0$ upward. If $X$ is below $A_0$, none is covered; if $X$ lies within the program, coverage stops at $X$; if $X$ exceeds $A_n$, coverage stops at $A_n$.

Consequently, the total recovery across the consecutive layers is the amount of the loss lying between the program’s bottom and top boundaries:

\sum_{j=1}^{n}R_j =\min\{\max(X-A_0,0),A_n-A_0\}

This result follows because the first covered band begins at $A_0$, each next band begins where the previous band ends, and the final band ends at $A_n$. There is no gap or overlap. The cedent’s retained amount, under the stated assumptions, is therefore

X-\sum_{j=1}^{n}R_j =\min(X,A_0)+\max(X-A_n,0)

The first term is the loss below the program’s attachment; the second is any loss above its top. Either term can be zero. This identity is a reconciliation check, not a substitute for checking individual layer recoveries and exhaustion.

Worked Example: A Loss Partly Uses the Top Layer

A hypothetical program has three consecutive layers: $300,000 excess of $400,000; $700,000 excess of $700,000; and $1,000,000 excess of $1,400,000. A single subject loss is $1,720,000. Calculate the recovery and exhaustion status of every layer and the cedent’s retained amount.

Method. List the bands: the first runs from $400,000 to $700,000; the second from $700,000 to $1,400,000; the third from $1,400,000 to $2,400,000.

First layer. The loss exceeds the $700,000 upper boundary, so the layer is exhausted. Its recovery is $\min(1{,}720{,}000-400{,}000,300{,}000)=300{,}000.$

Second layer. The loss also exceeds this layer’s $1,400,000 upper boundary. It is exhausted, with recovery $\min(1{,}720{,}000-700{,}000,700{,}000)=700{,}000.$

Third layer. The loss is between its $1,400,000 attachment and $2,400,000 upper boundary. It is partially used, with recovery $\min(1{,}720{,}000-1{,}400{,}000,1{,}000{,}000)=320{,}000.$

Total and check. Total recovery is $300,000 + $700,000 + $320,000 = $1,320,000. The cedent retains $1,720,000 − $1,320,000 = $400,000, exactly the amount below the $400,000 program attachment. The reconciliation is $1,320,000 + $400,000 = $1,720,000.

Common mistake avoided. Do not pay the full $1,000,000 limit of the third layer merely because the loss reaches that layer. Only $320,000 of the loss lies above its $1,400,000 attachment.

Losses Above the Program Top

A program’s top limit caps total recovery. If the subject loss exceeds the upper boundary of the highest layer, that excess remains with the cedent under the assumptions used here. It is not an additional recovery from the top layer. Keep this above-program amount separate from the amount below the first attachment.

Worked Example: All Layers Exhausted With Loss Above the Top

A hypothetical insurer buys three consecutive layers: $250,000 excess of $250,000; $750,000 excess of $500,000; and $1,000,000 excess of $1,250,000. A single subject loss is $2,800,000. Find all layer recoveries and the cedent’s retained amount.

Method. The layer boundaries are $250,000 to $500,000, $500,000 to $1,250,000, and $1,250,000 to $2,250,000. The loss exceeds the top boundary.

First layer. $R_1=\min(2{,}800{,}000-250{,}000,250{,}000)=250{,}000.$ It is exhausted.

Second layer. $R_2=\min(2{,}800{,}000-500{,}000,750{,}000)=750{,}000.$ It is exhausted.

Third layer. $R_3=\min(2{,}800{,}000-1{,}250{,}000,1{,}000{,}000)=1{,}000{,}000.$ It is exhausted.

Total and check. The total recovery is $250,000 + $750,000 + $1,000,000 = $2,000,000. The cedent retains $2,800,000 − $2,000,000 = $800,000. This consists of $250,000 below the program attachment and $550,000 above its $2,250,000 top: $250,000 + ($2,800,000 − $2,250,000) = $800,000. The check is $2,000,000 + $800,000 = $2,800,000.

Common mistake avoided. The top layer pays its $1,000,000 limit, not the $1,550,000 by which the loss exceeds that layer’s attachment. The layer limit caps its recovery.

Checks and Common Exam Traps

A reliable allocation has three checks: each recovery is between zero and its layer limit; no layer pays for loss outside its band; and total recovery plus the amount retained equals the subject loss. For consecutive layers, also check that each upper attachment equals the prior layer’s upper boundary.

Applying attachment to a remaining balance. Each layer’s formula uses the original subject loss $X$. Subtracting earlier recoveries before testing the next attachment shifts the loss into the wrong band.
Ignoring the layer limit. A layer’s recovery cannot exceed $L$, even if $X-A$ is larger than $L$.
Confusing entry with exhaustion. A loss just above attachment starts a recovery; it does not exhaust the layer. Exhaustion requires reaching the upper boundary.
Forgetting loss above the top. When the loss exceeds the highest boundary, the excess is retained unless other coverage is specified.
Assuming the layers are consecutive without checking. A gap or overlap changes the allocation. Confirm the stated boundaries before using the consecutive-layer total formula.

Key takeaway: For each layer, calculate $\min\{\max(X-A,0),L\}$ using the same subject loss $X$. Add the layer recoveries, classify exhaustion by comparing $X$ with each upper boundary, and reconcile total recovery with the cedent’s retained amount.

Check Your Understanding

For each problem, assume the stated layers apply to one subject loss, are consecutive where described, and have no other adjustments.

A single loss is $580,000. One layer is $250,000 excess of $400,000. Find its recovery and the cedent’s retained amount. (Answer: Recovery $180,000; retained $400,000.)
A program has layers of $300,000 excess of $200,000 and $500,000 excess of $500,000. A loss is $820,000. Find both recoveries and state which layer is exhausted. (Answer: Recoveries $300,000 and $320,000; first layer exhausted, second partially used; total recovery $620,000.)
Three consecutive layers are $400,000 excess of $400,000, $800,000 excess of $800,000, and $1,200,000 excess of $1,600,000. The loss is $3,050,000. Find each recovery and the retained amount, distinguishing below-program and above-program retention. (Answer: Recoveries $400,000, $800,000, and $1,200,000; retained $650,000, comprising $400,000 below attachment and $250,000 above the $2,800,000 top.)
A program has consecutive layers of $150,000 excess of $350,000, $450,000 excess of $500,000, and $700,000 excess of $950,000. A loss is $1,120,000. Calculate each recovery, identify exhausted layers, and find the retained amount. (Answer: Recoveries $150,000, $450,000, and $170,000; first two exhausted, third partially used; retained $350,000.)
A program has layers of $600,000 excess of $400,000, $1,000,000 excess of $1,000,000, and $1,500,000 excess of $2,000,000. A loss is exactly $2,000,000. Find the recoveries and explain the status of the third layer. (Answer: Recoveries $600,000, $1,000,000, and $0; the first two are exhausted, the third is unentered because the loss equals its attachment; retained $400,000.)

Understanding how one loss moves through a layered program supports broader risk-transfer analysis aligned with the CAS Exam 9 (Risk Management for Actuaries) syllabus.

What You'll Learn