Placement StructuresApplied

Reinsurance Tower

Q: Why do lower layers carry a much higher rate on line?

They are exposed to far more frequent losses — Rate on line tracks expected loss frequency. A working layer may be hit most years; the top layer may be hit once in fifty.

Q: Layer 2 above is placed at only 80%. A $30M event occurs. What does the cedent bear from that layer?

20% of the $5M falling in the layer = $1M — A $30M loss puts $5M into Layer 2 ($25M–$30M). With 80% placed, reinsurers pay $4M and the cedent co-reinsures $1M of its own layer.

Layering the programme

Cover stacked in layers above the retention, each priced and placed as its own contract.

Open the interactive version → definition, quiz, structure diagram and progress tracking

Definition

A tower (or programme) is the vertical stack of excess-of-loss layers a cedent buys above its retention. Each layer is a separate contract with its own limit, premium, reinstatement terms and panel of reinsurers. Layers are stacked because appetite is not uniform: the market that will write the frequency-exposed bottom layer is not the market that wants the remote top one.

Lower ("working") layers are hit often, so they carry a high rate on line — sometimes 20–40%. Top layers are remote and may cost 1–3%.
A layer placed at less than 100% leaves the cedent co-reinsuring the shortfall on its own account.
The top of the tower — retention plus all limits — is the number to compare against the modelled 1-in-250 loss.

Top of tower = Retention + Σ layer limits

Losses above this point are entirely unreinsured, whatever the programme cost.

Worked example

A tower of $15M xs $10M, $25M xs $25M, $50M xs $50M and $150M xs $100M gives $240M of cover above a $10M retention — the programme exhausts at $250M, and every dollar above that is the cedent's.

The structure diagram for this term is in the interactive version.

Scenario · figures in USD

Anatomy of a four-layer property catastrophe tower

Sentinel Regional retains $10M per event and buys four layers. Notice how the rate on line collapses as you climb.

Retention	$10M · uninsured
Layer 1 — $15M xs $10M	Premium $3.75M · ROL 25.0%
Layer 2 — $25M xs $25M	Premium $3.00M · ROL 12.0%
Layer 3 — $50M xs $50M	Premium $3.00M · ROL 6.0%
Layer 4 — $150M xs $100M	Premium $3.75M · ROL 2.5%
Total limit purchased	$240M
Top of tower	$250M
Total programme cost	$13.50M
Blended rate on line	5.63%

So whatThe bottom $15M of cover costs the same as the top $150M. That is the price of frequency, and it is why raising the retention is usually the cheapest lever a cedent has.

Check your understanding

Why do lower layers carry a much higher rate on line?

They are exposed to far more frequent losses. Rate on line tracks expected loss frequency. A working layer may be hit most years; the top layer may be hit once in fifty.

Layer 2 above is placed at only 80%. A $30M event occurs. What does the cedent bear from that layer?

20% of the $5M falling in the layer = $1M. A $30M loss puts $5M into Layer 2 ($25M–$30M). With 80% placed, reinsurers pay $4M and the cedent co-reinsures $1M of its own layer.

Word problem

Using Sentinel's tower above ($10M retention; layers of $15M xs $10M, $25M xs $25M, $50M xs $50M, $150M xs $100M), a windstorm produces a $180M ultimate net loss. Allocate it across the retention and each layer.

Show a hint

Fill the tower from the bottom up. Each layer absorbs its full limit before the next one is touched.

Reveal the worked answer

Retention absorbs the first $10M (loss now at $10M)
Layer 1 absorbs $10M → $25M = $15M (full limit)
Layer 2 absorbs $25M → $50M = $25M (full limit)
Layer 3 absorbs $50M → $100M = $50M (full limit)
Layer 4 absorbs $100M → $180M = $80M of its $150M limit
Check: 10 + 15 + 25 + 50 + 80 = $180M ✓

Cedent $10M; Layers 1–3 fully exhausted at $15M, $25M and $50M; Layer 4 pays $80M with $70M of limit still standing. Recoveries total $170M against a programme cost of $13.5M — and the cedent has now used three of its four layers, so reinstatement terms suddenly matter a great deal.