Modeled loss trigger
A mechanism running an event's physical parameters through a frozen model whose output triggers payment.
A modeled loss trigger works in two steps. First, an independent source records the event's actual physical parameters: track, intensity and footprint. Then they are fed into a catastrophe model and a reference portfolio, both frozen at issuance. The amount the model computes, not the loss actually suffered, determines payment. It strikes a balance between speed and relevance. An indemnity trigger tracks the sponsor's losses but settles over years. A parametric trigger settles in days but diverges widely from actual damage. Modeled loss settles in weeks while still reflecting portfolio structure. Basis risk remains, but of a different kind: it now comes from the gap between the frozen model and reality, no longer from the difference between sponsor and market. A model that ages badly over the note's life favors or penalizes one side, which is why these structures are written short.
A 2026 earthquake note defines a reference portfolio of 12,400 sites and a model frozen at the version in force at issuance. A quake occurs: recorded magnitude and depth are fed into the model, which computes 143 million euros of losses on the reference portfolio. The note pays on that basis fourteen weeks after the event, while the actual indemnity run-off will not close until 2032.
Follow the research
One letter a week. First, the week's underwriting note, when one is published: a dated, argued decision, signed by Paul Bernardski. Then what the observatory published on cyber risk, political risk and insurance. No account, no tracking.
No spam, one-click unsubscribe, no data ever sold.