The gap between the reserve carried at a year end and the cost actually paid afterward, the only public judge of how prudently a book was reserved.
A run-off surplus arises where the claims of a cohort finally cost less than what was set aside; a run-off deficit is the reverse. The quantity is mechanical, but reading it is delicate: a steady surplus can signal prudent and solid management, or on the contrary systematic over-reserving that artificially smooths results from year to year. A repeated deficit leaves no room for doubt, since it means the firm published for years results that did not exist. This is why regulators made public the triangles of paid and reserved claims by accident year, which expose ten years of history and let an outside reader judge without access to the files. The problem solved is verifying an estimate that is by nature unverifiable when made, and becomes checkable only in hindsight.
The solvency and financial condition report, which every European insurer has published annually since 2017, contains quantitative template S.19.01: ten accident years down the rows, ten development years across the columns, for payments and for reserves. It is the only public data allowing anyone, without inside information, to measure whether a firm has systematically over-reserved or under-reserved.
résultat de liquidation, dérive des provisions