BootstrapBornFerg: Bootstrap

In ArnaudBu/ReservingLad: Reserving for Non Life Insurance

Description

BootstrapBornFerg computes the bootstraped distribution for IBNR on a Bornhuetter Fergusson method.

Usage

BootstrapBornFerg(triangle, ultimateClaims, nBoot = 1000, weight = NA)

Arguments

triangle	Cumulated triangle as a matrix. The matrix should be square
ultimateClaims	Ultimate claims by accident year.
nBoot	Number of samples. Default value equals to 1000.
weight	Boolean matrix the same size of the triangle to tell if the value is to be considered: 1 for yes, 0 for no. First column is not considered

Value

A list containing the following objects:

• triangleList: the list of bootsraped triangles

• ibnrByAccidentYear: a data frame containing ibnrs by accident year

• ibnr: a vector of the total ibnr values)

• predictionError: estimation of the prediction error on IBNR

Examples

ultimateClaims <- c(3901463,5433719,5378826,5297906,4858200,5111171,5660771,6784799,5642266,4969825)
bbf <- BootstrapBornFerg(triangleExampleEngland, ultimateClaims, 1000)

ArnaudBu/ReservingLad documentation built on Sept. 21, 2021, 1:19 p.m.