accountForZeroPayments: A function to take a triangle estimated without considering...

In BALD: Robust Loss Development Using MCMC

Description

A function to take a triangle estimated without considering zero payments, and account for the possibility of zero payments.

Usage

accountForZeroPayments(object, burnIn=1000, nAddapt=1000)

Arguments

object	The object containing the triangle estimated without accounting for zero payments.
burnIn	An integer to represent the number of initial MCMC iterations to be discarded. (The adaptive phase (nAddapt) is not considered part of burnIn.)
nAddapt	The length of the adaptive phase for the MCMC algorithm. (Default is trunc(burnIn/4)+1.)

Details

As incremental payments are modeled on the log scale, zero payments (and negative payments) are treated as missing values. So, without somehow accounting for zero payments, the estimated payments would be overstated. Zero payments are accounted for by weighting the predicted payment (given that the payment is greater than zero) with the probability that this payment is zero. (Negative payments are not (currently) accounted for.) Currently the trajectory for this probably follows a gompertz curve and is constant across exposure years. This is currently implemented as a function but may be switched to a method. See vignette('BALD').

Examples

rm(list=ls())
library(BALD)
data(CumulativeAutoBodilyInjuryTriangle)
CumulativeAutoBodilyInjuryTriangle <- as.matrix(CumulativeAutoBodilyInjuryTriangle)
sample.col <- (dim(CumulativeAutoBodilyInjuryTriangle)[2] - 6:0)
print(decumulate(CumulativeAutoBodilyInjuryTriangle)[1:7, sample.col])
data(HPCE)
HPCE <- as.matrix(HPCE)[,1]
HPCE.rate <- HPCE[-1] / HPCE[-length(HPCE)] - 1
print(HPCE.rate[(-10):0 + length(HPCE.rate)])
HPCE.years <- as.integer(names(HPCE.rate))
max.exp.year <- max(as.integer(
dimnames(CumulativeAutoBodilyInjuryTriangle)[[1]]))
years.to.keep <- HPCE.years <=  max.exp.year + 3
HPCE.rate <- HPCE.rate[years.to.keep]
break.model.input <- makeBreakAnnualInput(
cumulative.payments = CumulativeAutoBodilyInjuryTriangle,
stoch.inflation.weight = 1,
non.stoch.inflation.weight = 0,
stoch.inflation.rate = HPCE.rate,
first.year.in.new.regime = c(1986, 1987),
prior.for.first.year.in.new.regime=c(2,1),
exp.year.type = 'ay',
extra.dev.years = 5,
use.skew.t = TRUE,
bound.for.skewness.parameter=5)
## Not run: 
break.model.output <- runLossDevModel(
break.model.input,
burnIn=30.0E+3,
sampleSize=30.0E+3,
thin=10)
break.model.output.w.zeros <- accountForZeroPayments(break.model.output)

## End(Not run)

BALD documentation built on May 2, 2019, 6:51 a.m.