01.06.2015 Views

Actuarial Modelling of Claim Counts Risk Classification, Credibility ...

Actuarial Modelling of Claim Counts Risk Classification, Credibility ...

Actuarial Modelling of Claim Counts Risk Classification, Credibility ...

SHOW MORE
SHOW LESS

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

170 <strong>Actuarial</strong> <strong>Modelling</strong> <strong>of</strong> <strong>Claim</strong> <strong>Counts</strong><br />

future (the level for year t + 1) depends on the present (the level for year t and the number<br />

<strong>of</strong> accidents reported during year t) and not on the past (the complete claim history and<br />

the levels occupied during years 1 2t− 1). Sometimes, fictitious levels have to be<br />

introduced in order to meet this memoryless property. Indeed, in some bonus-malus systems,<br />

policyholders occupying high levels are sent to the starting class after a few years without<br />

claims. This issue will be addressed in Section 4.7.<br />

4.2.3 Trajectory<br />

New drivers start in level l 0 <strong>of</strong> the scale. Note that experienced drivers arriving in the<br />

portfolio are not necessarily placed in level l 0 , but in a level corresponding to their claim<br />

history or to the level occupied in the bonus-malus scale used by a competitor. This problem<br />

will be dealt with in Section 4.8.<br />

The trajectory <strong>of</strong> the policyholder in the bonus-malus scale is modelled by a sequence<br />

L 1 L 2 <strong>of</strong> random variables valued in 0 1s, such that L k is the level occupied<br />

during the k + 1th year, i.e. during the time interval k k + 1. Since movements in the<br />

scale occur once a year (at policy anniversary), the policyholder occupies level L k from<br />

time k until time k + 1. Once the number N k <strong>of</strong> claims reported by the policyholder during<br />

k − 1k is known, this information is used to reevaluate the position <strong>of</strong> the driver in the<br />

scale. We supplement the sequence <strong>of</strong> the L k s with L 0 = l 0 .<br />

The L k s obviously depend on the past numbers <strong>of</strong> claims N 1 N 2 N k reported by<br />

the policyholder. If we denote as ‘pen’ the penalty induced by each claim (expressed as a<br />

number <strong>of</strong> levels), then the L k s obey the recursion<br />

{ maxLk−1 − 1 0 if N<br />

L k =<br />

k = 0<br />

minL k−1 + N k × pens if N k ≥ 1<br />

{<br />

= max min { L k−1 − 1 − I k + N k × pens } }<br />

0<br />

where<br />

{ 1ifNk ≥ 1<br />

I k =<br />

0 otherwise<br />

indicates whether at least one claim has been reported in year k. This is an example <strong>of</strong> a<br />

stochastic recursive equation. This representation <strong>of</strong> the L k s clearly shows that the future<br />

trajectory <strong>of</strong> the policyholder in the scale is independent <strong>of</strong> the levels occupied in the past,<br />

provided that the present level is given. This conditional independence property is at the<br />

heart <strong>of</strong> Markov models.<br />

The stochastic recursive equations given above assume that the bonus is lost in case at<br />

least one claim is filed with the company. In some cases (like the former compulsory Belgian<br />

bonus-malus scale), the bonus is granted in any case. The L k s then obey the recursion<br />

{<br />

L k = max min { L k−1 − 1 + N k × pens } }<br />

0 <br />

This means that the first claim is penalized by pen−1 levels, and the subsequent ones by<br />

pen levels.

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!