Match a reinsurance tower — match

One collective model (a frequency and a piecewise Pareto severity) that reproduces the expected loss of every layer of a tower (Riegel 2018, Matching Algorithm 2).

Usage

match_tower(
  attachments,
  layer_losses,
  frequencies = NULL,
  rule = c("minimize", "midpoint")
)

Arguments

attachments: Increasing attachment points; layer i runs to the next one, the last is unlimited.
layer_losses: Expected loss a year of each layer.
frequencies: Expected losses a year above each attachment point, NA where to derive them, or NULL to derive all.
rule: "minimize" (make the two alphas in each layer as close as possible) or "midpoint".

Value

A tower_model.

Examples

m <- match_tower(c(1000, 1500, 2000, 2500, 3000, 5000, 10000),
                 c(100, 90, 50, 40, 100, 50, 50),
                 frequencies = c(0.25, rep(NA, 6)))
layer(m, 500, 1500)
#> [1] 90