Catastrophe Protection in Layered Programs

How Layered Catastrophe Protection Responds

A catastrophe program may protect an insurer against a large loss from a single event while limiting how much the reinsurer pays across the contract year. To estimate the protection, the actuary must distinguish the occurrence loss from the treaty’s attachment point, per-occurrence limit, and any annual cap on recoveries. Reinstatements can change the amount of per-occurrence capacity available after a layer has been exhausted.

The calculations in this tutorial treat reported occurrence losses and contract terms as known hypothetical amounts. They are deterministic examples: no probability distribution or mortality assumption is used, and no interest or discounting is applied. In practice, the loss definition, event grouping, currency, contract period, and order of coverage calculations must follow the actual treaty wording.

Definition: For an occurrence with covered loss $L$, a layer attaching at retention $R$ with per-occurrence limit $U$ pays the lesser of the amount above $R$ and $U$. Its recovery before any aggregate cap or capacity restriction is $\min(\max(L-R,0),U)$. The insurer’s net loss for that occurrence is $L$ minus the recovery, assuming $L$ is already the treaty-defined covered loss.

A layer described as $U$ excess of $R$ attaches after the insurer has borne $R$ of covered occurrence loss; it does not mean that the reinsurer pays $U$ for every occurrence. If the occurrence loss is at or below $R$, there is no recovery. If it exceeds $R$ by less than $U$, the recovery is the excess. If it exceeds $R+U$, the layer is exhausted and pays $U$, subject to other treaty restrictions.

C(L)=\min\bigl(\max(L-R,0),U\bigr)

Here $C(L)$ is the layer’s calculated per-occurrence recovery before applying an annual aggregate cap or a restriction on remaining treaty capacity. A contract may also specify an annual aggregate limit, meaning a maximum total payment over the contract period. If $A$ is the aggregate limit and $P$ is the amount already paid under that limit, the remaining aggregate capacity is $\max(A-P,0)$. The current recovery cannot exceed that remaining capacity.

C_{\text{pay}}=\min\bigl(C(L),\max(A-P,0)\bigr)

This expression assumes the annual aggregate cap applies to the layer’s recoveries and that $P$ records the amount already counted against that cap. If the contract instead has separate aggregates, special loss definitions, or other limits, the calculation must reflect those terms. A cap is not another attachment point: it limits the total amount paid, rather than changing the occurrence loss at which coverage begins.

A Reliable Calculation Sequence

For a layered catastrophe program, calculate the amount that would be payable under the occurrence terms first, then apply any cap or capacity restriction. Keep the loss basis consistent throughout: do not mix gross policy loss with treaty-defined loss if the contract specifies a different basis.

Establish the covered occurrence loss.
Use the treaty’s definition of covered loss and occurrence. Aggregate claims into an occurrence only as the contract requires.

Calculate the layer’s uncapped recovery.
Subtract the retention, floor the result at zero, and cap it at the per-occurrence limit.

Apply remaining aggregate or available capacity.
Limit the calculated recovery to the amount still payable under the stated annual aggregate and reinstatement terms.

Reconcile the result.
For the covered loss basis, net loss equals covered loss less the recovery actually payable.

Why the Layer Formula Works

The formula follows directly from the layer’s payment structure. The loss amount above the retention is $L-R$, but a loss below the retention cannot produce a negative recovery. Thus the covered amount reaching the layer is $\max(L-R,0)$. The treaty cannot pay more than its per-occurrence limit $U$, so the payable amount before other restrictions is the lesser of that excess and $U$.

\text{amount above retention}=\max(L-R,0) $$ $$ \text{layer recovery}=\min\bigl(\text{amount above retention},U\bigr)

For example, when $L \le R$, the first expression is zero, so the layer pays zero. When $R < L < R+U$, the excess is positive but below the limit, so the layer pays $L-R$. When $L \ge R+U$, the excess is at least $U$, so payment is capped at $U$. These cases establish the formula for losses below, within, and above the layer.

Worked Example: Combining Claims Into One Occurrence

A hypothetical insurer has a catastrophe layer of $ \$5{,}000{,}000 $ excess of $ \$8{,}000{,}000 $, with no annual aggregate restriction or reinstatement issue. A single covered event produces three treaty-defined claim amounts: $ \$4{,}200{,}000 $, $ \$3{,}600{,}000 $, and $ \$5{,}450{,}000 $. Find the event recovery and the insurer’s net event loss. Use nominal dollars with no discounting; the stated claim amounts are fixed, so no probability distribution is assumed.

Method. First combine the claims into the covered loss for the single occurrence:

$$ L=4{,}200{,}000+3{,}600{,}000+5{,}450{,}000 =13{,}250{,}000 $$

The amount above the $ \$8{,}000{,}000 $ retention is $ \$13{,}250{,}000-\$8{,}000{,}000=\$5{,}250{,}000 $. The per-occurrence limit is $ \$5{,}000{,}000 $, so the recovery is:

C(L)=\min(5{,}250{,}000,5{,}000{,}000)=\$5{,}000{,}000

The insurer retains $ \$13{,}250{,}000-\$5{,}000{,}000=\$8{,}250{,}000 $ of this event loss. The check is that the retention plus the layer payment is $ \$8{,}000{,}000+\$5{,}000{,}000=\$13{,}000{,}000 $; the remaining $ \$250{,}000 $ is above the top of the layer and is retained.

Common mistake avoided: Applying the $ \$8{,}000{,}000 $ retention separately to each claim would ignore the stated single-occurrence basis. The treaty applies its occurrence terms to the combined $ \$13{,}250{,}000 $ loss, not three independent retentions.

Annual Aggregate Limits and Remaining Capacity

An annual aggregate limit can reduce a recovery that otherwise fits within a per-occurrence layer. The key quantity is the aggregate limit still available immediately before the current recovery. If the annual cap is $A$ and prior payments counted against it total $P$, then at most $A-P$ remains when $P < A$, and none remains once $P\ge A$. The current payment is the smaller of the per-occurrence recovery and that remaining amount.

Worked Example: Applying an Annual Aggregate Cap

A hypothetical treaty provides $ \$4{,}000{,}000 $ per occurrence excess of $ \$6{,}000{,}000 $, subject to a $ \$5{,}000{,}000 $ annual aggregate limit. Earlier in the year, $ \$2{,}000{,}000 $ has been paid and counted against the aggregate. A new covered occurrence has a treaty-defined loss of $ \$12{,}000{,}000 $. Determine the recovery now payable and the insurer’s net loss for this occurrence. Amounts are nominal, with no discounting; there is no stochastic loss assumption.

Method. The current occurrence’s amount above the retention is $ \$12{,}000{,}000-\$6{,}000{,}000=\$6{,}000{,}000 $. Applying the per-occurrence limit gives:

C(L)=\min(6{,}000{,}000,4{,}000{,}000)=\$4{,}000{,}000

The annual aggregate has $ \$5{,}000{,}000-\$2{,}000{,}000=\$3{,}000{,}000 $ remaining. Therefore, the actual payment is the lesser of $ \$4{,}000{,}000 $ and $ \$3{,}000{,}000 $:

C_{\text{pay}}=\min(4{,}000{,}000,3{,}000{,}000)=\$3{,}000{,}000

The insurer’s net loss for this occurrence is $ \$12{,}000{,}000-\$3{,}000{,}000=\$9{,}000{,}000 $. The aggregate check is $ \$2{,}000{,}000+\$3{,}000{,}000=\$5{,}000{,}000 $, exactly the annual cap.

Common mistake avoided: Paying the full $ \$4{,}000{,}000 $ layer recovery would exceed the annual aggregate: prior payments plus that recovery would be $ \$6{,}000{,}000 $. The annual cap restricts payment to the $ \$3{,}000{,}000 $ still available.

Reinstatements and Exhausted Capacity

A reinstatement provision can restore some or all of a layer’s limit after it has been used, subject to its wording. To calculate later recoveries, track the limit available at each point in the treaty period. A reinstatement is not automatically an unlimited additional layer: the number and amount of reinstatements, when restored capacity becomes available, and any aggregate maximum are contract terms.

For the following illustration, assume one full reinstatement of the per-occurrence limit. It becomes available immediately after the original limit is fully exhausted. The example has no separate annual aggregate cap, no other capacity restriction, and no reinstatement premium calculation. Losses are nominal fixed amounts; no interest, mortality, or probability distribution is assumed.

Worked Example: Tracking One Full Reinstatement

A hypothetical catastrophe treaty covers $ \$3{,}000{,}000 $ excess of $ \$2{,}000{,}000 $ per occurrence, with one full reinstatement. Three successive covered occurrences have losses of $ \$6{,}000{,}000 $, $ \$4{,}000{,}000 $, and $ \$3{,}000{,}000 $. Assume the original limit is available for the first occurrence, the full reinstated limit becomes available after the original limit is exhausted, and no other annual cap applies. Calculate each recovery and the total recovery.

Method. For the first occurrence, the amount above retention is $ \$6{,}000{,}000-\$2{,}000{,}000=\$4{,}000{,}000 $. The original $ \$3{,}000{,}000 $ limit caps the recovery at $ \$3{,}000{,}000 $. The original capacity is now exhausted, so the full reinstatement restores $ \$3{,}000{,}000 $ of capacity.

For the second occurrence, the amount above retention is $ \$4{,}000{,}000-\$2{,}000{,}000=\$2{,}000{,}000 $. This is below the restored limit, so the recovery is $ \$2{,}000{,}000 $. The restored capacity remaining is $ \$3{,}000{,}000-\$2{,}000{,}000=\$1{,}000{,}000 $.

For the third occurrence, the amount above retention is $ \$3{,}000{,}000-\$2{,}000{,}000=\$1{,}000{,}000 $. That equals the remaining capacity, so the recovery is $ \$1{,}000{,}000 $. Total recovery is:

3{,}000{,}000+2{,}000{,}000+1{,}000{,}000=\$6{,}000{,}000

The check is that the treaty has paid no more than the original $ \$3{,}000{,}000 $ limit plus one $ \$3{,}000{,}000 $ reinstatement. The third occurrence receives only the $ \$1{,}000{,}000 $ still available, not the full amount that its layer calculation would otherwise suggest.

Common mistake avoided: Treating the reinstatement as a fresh, unlimited $ \$3{,}000{,}000 $ limit for every later occurrence would produce excess recoveries. Only one restored limit is available in this example, and its remaining balance must be tracked.

Exam Traps and Practical Checks

The most reliable check is to reconcile covered loss, recovery, and net loss for each occurrence, while separately tracking any aggregate or reinstated capacity. The following errors can lead to plausible-looking but incorrect results:

Applying a retention per claim instead of per occurrence. First determine the treaty-defined occurrence loss; then apply the occurrence retention once to that loss.
Ignoring the top of the layer. A loss above $R+U$ does not produce a recovery greater than $U$, unless additional, separately specified capacity applies.
Applying the aggregate cap before calculating the layer recovery. Calculate the per-occurrence amount first, then restrict it to the remaining aggregate capacity.
Forgetting prior aggregate payments. The cap limits payments over the stated period; the full annual cap is not available anew for each occurrence.
Assuming reinstatements without checking terms. Restored capacity depends on the number, amount, and availability conditions in the treaty.

These calculations matter because a gross catastrophe estimate does not by itself indicate the insurer’s retained loss. A layer may be below attachment, exhausted at its limit, constrained by an annual aggregate, or operating on reinstated capacity. Accurate estimates require the covered loss basis and the contract restrictions to be applied in a consistent order.

Key takeaway: Calculate each occurrence’s layer recovery from its covered loss, retention, and per-occurrence limit. Then apply any remaining aggregate cap or available reinstated capacity. The actual recovery cannot exceed any applicable restriction, and net covered loss equals covered loss minus the recovery actually payable.

Check Your Understanding

For each hypothetical problem, calculate the requested amount using only the stated treaty terms.

A layer is $ \$2{,}500{,}000 $ excess of $ \$4{,}000{,}000 $, with no aggregate restriction. One occurrence has covered loss of $ \$5{,}800{,}000 $. What is the recovery? (Answer: $ \$1{,}800{,}000 $)
A layer is $ \$3{,}000{,}000 $ excess of $ \$7{,}000{,}000 $. One occurrence has loss of $ \$13{,}500{,}000 $, and the annual aggregate has $ \$2{,}000{,}000 $ remaining. Find the actual recovery and net loss. (Answer: Recovery $ \$2{,}000{,}000 $; net loss $ \$11{,}500{,}000 $)
A $ \$4{,}000{,}000 $ layer excess of $ \$5{,}000{,}000 $ has an annual aggregate cap of $ \$6{,}000{,}000 $. Prior payments are $ \$4{,}500{,}000 $. A new occurrence has covered loss of $ \$10{,}000{,}000 $. How much is payable now, and how much aggregate remains afterward? (Answer: Payable $ \$1{,}500{,}000 $; aggregate remaining $ \$0 $)
A $ \$2{,}000{,}000 $ layer excess of $ \$1{,}000{,}000 $ has one full reinstatement, with no separate aggregate cap. Successive occurrence losses are $ \$4{,}000{,}000 $, $ \$2{,}500{,}000 $, and $ \$3{,}000{,}000 $. Assume the reinstatement is available after the first limit is exhausted. Find the three recoveries and their total. (Answer: $ \$2{,}000{,}000 $, $ \$1{,}500{,}000 $, and $ \$500{,}000 $; total $ \$4{,}000{,}000 $)
A hypothetical program has a $ \$3{,}000{,}000 $ layer excess of $ \$6{,}000{,}000 $, a $ \$4{,}000{,}000 $ annual aggregate, and $ \$1{,}250{,}000 $ of prior aggregate payments. A new occurrence has covered loss $ \$12{,}000{,}000 $. Calculate the recovery now payable and the insurer’s net loss for that occurrence. (Answer: Recovery $ \$2{,}750{,}000 $; net loss $ \$9{,}250{,}000 $)

What You'll Learn