layer, attachment point, and limit

Think of a tall building of risk. The ground floors — the small, frequent losses — are kept by the cedant. Higher up, each storey is a band of larger possible losses, and a reinsurer can be hired to cover one particular storey: it pays losses that fall inside that band and nothing outside it. These storeys are layers, and they are the vocabulary in which most non-proportional reinsurance is described and traded.

Three numbers define a layer. The attachment point is the floor — the loss level at which the layer begins to respond; below it the layer pays nothing. The limit is the height of the layer — the most it will pay. Reinsurers write a layer as 'limit excess of attachment', so '5 million excess of 5 million' is a layer that begins paying once a loss exceeds 5 million and pays up to a further 5 million (covering the band from 5 to 10 million). A 12 million loss against that layer pays its full 5 million limit; the cedant keeps the first 5 million and anything above 10 million. The exhaustion point, where the layer is used up, is the attachment plus the limit.

Layers matter because they are the building blocks of an entire reinsurance program. Risk is sliced into stacked layers, and different reinsurers buy different storeys, often at very different prices: low layers attach close to expected losses and are expensive per dollar of cover, while high layers respond only to true catastrophes and are cheap per dollar but enormous in total limit. Picking where to set each attachment and limit is a core actuarial exercise balancing protection against cost. A frequent confusion: the attachment point of a reinsurance layer is the reinsurer's starting line, not the policyholder's deductible — it sits far above any single customer's claim and concerns the insurer's own losses.