Learn R Programming

ProfileLadder (version 0.2.2)

predict.profileLadder: One-year-ahead prediction based on PARALLAX, REACT, or MACRAME

Description

The function provides a one-year-ahead prediction for a given run-off triangle (i.e., estimating a new running diagonal based on some nonparametric, functional-based prediction algorithm, PARALLAX, REACT, or MACRAME).

Usage

# S3 method for profileLadder
predict(object, ...)

Value

An object of the class profilePredict which is a list with the following elements:

reserve

The overall predicted amount of the next year payments---the new running diagonal in an extended (cumulative) run-off triangle minus the last observed diagonal (in the cumulative triangle again)

methods

Provides the information about the underlying nonparametric (functional-based) method that is used for the 1-year-ahead prediction (PRALLAX, REACT, or MACRAME)

newDiagonal

Numeric vector representing the prediction of the new running diagonal (ordered in a way that first value corresponds to the payment for the last occurrence period (typically the largest amount) and the last value corresponds with the next-year payment for the first occurrence periods---typically the smallest amount)

extTriangle

The extended run-off triangle---the R object of the class triangle that also contains the new predicted diagonal---the 1-year-ahead prediction

Arguments

object

an object of the class profileLadder which is the output from the parallelReserve() function or the mcReserve function

...

further arguments passed to predict()

See Also

parallelReserve(), mcReserve()

Examples

Run this code
## full run-off triangle complettion
parallelReserve(CameronMutual)

## one-year-ahead prediction based on PARALLAX
predict(parallelReserve(CameronMutual))

Run the code above in your browser using DataLab