Scribd
Upload
31 views

Claims Triangle - R: Submitted by Ishan Bandyopadhyay

This document discusses various statistical methods for estimating outstanding claims reserves in general insurance, including the chain-ladder method. It describes the key steps to applying the chain-ladder method using a run-off triangle, including compiling data, calculating age-to-age factors, selecting development factors, and projecting ultimate claims. It also summarizes several extensions of the chain-ladder method, such as the Mack chain-ladder model, Munich chain-ladder, bootstrap chain-ladder, and Clark's LDF method.

Uploaded by

Agni Banerjee
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
31 views

Claims Triangle - R: Submitted by Ishan Bandyopadhyay

This document discusses various statistical methods for estimating outstanding claims reserves in general insurance, including the chain-ladder method. It describes the key steps to applying the chain-ladder method using a run-off triangle, including compiling data, calculating age-to-age factors, selecting development factors, and projecting ultimate claims. It also summarizes several extensions of the chain-ladder method, such as the Mack chain-ladder model, Munich chain-ladder, bootstrap chain-ladder, and Clark's LDF method.

Uploaded by

Agni Banerjee
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 23

Claims triangle

-R Submitted by
Ishan Bandyopadhyay
Insurers don’t know the upfront cost for their
service, but rely on historical data analysis and
judgement to predict a sustainable price for
their offering.
The claims payment process can take years or
even decades. Therefore, often not even the
Introduction delivery date of their product is known to
insurers.
It should come as no surprise that the biggest
item on the liabilities side of an insurer’s
balance sheet is often the provision or
reserves for future claims payments.
The ChainLadder provides various statistical methods
which are typically used for the estimation of outstanding
claims reserves in general insurance, including those to
estimate the claims development results.

The key idea is to see the chain-ladder algorithm as a


special form of a weighted linear regression through the
“ChainLadder” origin, applied to each development period.

The chain ladder method calculates incurred but not


reported (IBNR) loss estimates, using run-off triangles of
paid losses and incurred losses, representing the sum of
paid losses and case reserves.
Compile Compile claims data in a development triangle

Calculate Calculate age-to-age factors

Calculate Calculate averages of the age-to-age factors

Steps to apply Select Select claim development factors

“ChainLadder”
Select Select tail factor

Calculate Calculate cumulative claim development factors

Project Project ultimate claims


R and
Databases
Loss triangle
4
25000

5 5
4
4 3 3
5 3
4
3 1 1 1
1
15000

3 1 2 2
5
4 6 2 2
RAA

3 2
1
8 6
6 7
1
4 7
1 2
5
3
1
8
6
5000

4
1 9 2
2
7
3
9
0
6
8
5
7
0

2 4 6 8 10

dev. period
Claims Development by origin period

2 4 6 8 10 2 4 6 8 10

1981 1982 1983 1984 1985


25000
20000
15000
10000
5000
0
1986 1987 1988 1989 1990
25000
20000
15000
10000
5000
0

2 4 6 8 10 2 4 6 8 10 2 4 6 8 10

dev. period
Loss triangle - Incremental
Log-linear extrapolation of age-to-age factors

0
-1
Log-linear

log(f - 1)

-2
model

-3
-4

2 4 6 8

dev.period
LDF method
• Thomas Mack published in 1993 a
method which estimates the standard
errors of the chain-ladder forecast
without assuming a distribution.
Mack • The Mack chain-ladder model gives an
unbiased estimator for IBNR (Incurred
ChainLadder But Not Reported) claims.
• The Mack chain-ladder model can be
regarded as a weighted linear
regression through the origin for each
development period.
Mack triangle
Munich ChainLadder

• Munich chain-ladder is a
reserving method that reduces
the gap between IBNR
projections based on paid losses
and IBNR projections based on
incurred losses.
• The Munich chain-ladder method
uses correlations between paid
and incurred losses of the
historical data into the projection
for the future.
• The Boot ChainLadder function uses a two-
stage bootstrapping/simulation approach
following the paper by England and Verrall.
• In the first stage an ordinary chain-ladder
methods is applied to the cumulative claims
triangle.
Bootstrap • In the second stage we simulate the process
error with the bootstrap value as the mean and
ChainLadder using the process distribution assumed.
• The set of reserves obtained in this way forms
the predictive distribution, from which
summary statistics such as mean, prediction
error or quantiles can be derived.
BootChainLadder
Quantiles of
Bootstrap
IBNR
• The LDF method assumes that the ultimate
losses in each origin period are separate
and unrelated. The goal of the method,
therefore, is to estimate parameters for
the ultimate losses and for the growth
curve in order to maximize the likelihood
of having observed the data in the triangle.
Clark’s • The CapeCod method assumes that the
apriori expected ultimate losses in each
methods origin year are the product of earned
premium that year and a theoretical loss
ratio. The CapeCod method, therefore,
need estimate potentially far fewer
parameters: for the growth function and
for the theoretical loss ratio.
Clark’s LDF method
GENERALISED LINEAR A PAID INCURRED CHAIN
MODELS [GLM] MODEL.

Future Scope

A MULTIVARIATE
CHAINLADDER APPROACH

You might also like