Non-Proportional CoversAdvanced

Index / Stability Clause

Keeping a retention's real value stable

A mechanism that adjusts a long-tail layer's attachment point and limit for inflation, so a retention set today still means the same thing after a decade of claims development.

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

Definition

An index clause (or stability clause) adjusts the attachment point and limit of a long-tail excess-of-loss layer in line with an agreed inflation index — often a wage or claims-cost index — between inception and the date a claim is finally settled. Without it, years of inflation on a slow-developing casualty claim erode the real value of a retention fixed in nominal dollars at placement, silently shifting loss from the cedent to the reinsurer as the layer effectively drops lower and lower against inflated claims.

It exists specifically for long-tail lines — casualty, medical malpractice, workers' compensation — where a claim can take a decade or more to settle, unlike property lines where the attachment point and the loss are fixed within the same year.
Both the attachment point and the limit are typically indexed together, so the layer's real width stays constant even as its nominal dollar boundaries both rise.
Clauses commonly include a threshold — index adjustment only applies once cumulative inflation since inception exceeds an agreed percentage — so the mechanism does not fire on ordinary year-to-year noise.

Indexed boundary = Original boundary × (Index at settlement ÷ Index at inception)

Applied to both the attachment point and the limit, so the layer keeps the same real width throughout the claim's development.

Worked example

A casualty layer of $10M xs $5M is placed when the wage index stands at 100. The claim settles eight years later, when the index has risen to 135. Both boundaries index up by 35%: the layer becomes $13.5M xs $6.75M in nominal terms, protecting the same real slice of the loss.

Scenario · figures in USD

The same real layer, a decade apart in nominal dollars

Granite Casualty places a $10M xs $5M per-claim layer on a long-tail liability treaty, indexed to a wage index standing at 100 at inception. A severe bodily injury claim, reserved at inception, finally settles nine years later when the index has risen to 142.

Original layer (index = 100 at inception)	$10M xs $5M
Index at settlement	142
Index ratio (142 ÷ 100)	1.42
Indexed attachment point ($5M × 1.42)	$7.10M
Indexed limit ($10M × 1.42)	$14.20M
Indexed layer	$14.20M xs $7.10M
Claim settles for	$19M
Recovery under the indexed layer	$11.90M
Recovery had the layer never been indexed	$10.00M

So whatIndexing did not just move the boundaries — it moved money. The cedent recovers $1.90M more under the indexed layer than the original nominal boundaries would have allowed, because the retention's real value was preserved rather than eroded by nine years of inflation.

Check your understanding

Why are index clauses mainly used on long-tail lines like casualty rather than property?

Long-tail claims can take many years to settle, during which inflation erodes a fixed nominal retention's real value. A property claim is fixed and settled within the same year as the loss, so there is little inflation gap to correct. A casualty claim developing over a decade is exactly where the erosion happens.

An index clause raises a layer's attachment point from $5M to $7.10M after nine years of inflation. Why doesn't this necessarily reduce the cedent's recovery on an inflated claim?

The limit rises by the same factor, so the whole layer is wider in nominal terms and can absorb more of an inflated claim. Both boundaries move together. A higher attachment point sounds like less cover, but a proportionally higher limit widens the layer in nominal terms by exactly the same factor.

Word problem

A long-tail layer of $8M xs $4M is placed when an agreed wage index stands at 110. A claim settles eleven years later when the index has risen to 176. The claim settles for $16M. Calculate the indexed attachment point, the indexed limit, and the recovery under the indexed layer — then compare it with what the original, unindexed boundaries would have recovered.

Show a hint

Find the index ratio first and apply it to both boundaries before working out the recovery either way.

Reveal the worked answer

Index ratio = 176 ÷ 110 = 1.60
Indexed attachment point = $4M × 1.60 = $6.4M
Indexed limit = $8M × 1.60 = $12.8M
Indexed layer covers $6.4M to $19.2M; recovery = $16M − $6.4M = $9.6M (within the $12.8M limit)
Unindexed layer would have covered $4M to $12M; recovery formula gives $16M − $4M = $12M, but is capped at the original $8M limit, giving $8M

The indexed layer recovers $9.6M against $8M under the original, unindexed boundaries — $1.6M more, because the limit widened by the same 1.60 factor as the attachment point. A higher attachment point sounds like less cover; indexing both boundaries together is what keeps that intuition from being true.