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

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

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

4.4.3 Dufresne Algorithm<br />

Dufresne (1988,1995) proposed a simple and efficient iterative algorithm for deriving <br />

provided that the driver goes one level down if no claims are filed to the company, and goes<br />

n×pen levels up if n claims are reported to the insurer. Then, the move at the end <strong>of</strong> year k<br />

can be modelled as<br />

{ −1 if no claims<br />

k+1 =<br />

n × pen if n claims<br />

If the annual numbers <strong>of</strong> claims N 1 N 2 are independent and oi distributed then the<br />

sequence 1 2 is made up <strong>of</strong> independent and identically distributed random<br />

variables, with common probability mass function<br />

Pr k+1 =−1 = PrN k+1 = 0 = exp−<br />

Pr k+1 = n × pen = PrN k+1 = n = exp− n<br />

n!<br />

Pr k+1 = = 0 otherwise<br />

The level L k+1 can then be represented as<br />

for n = 1 2<br />

⎧<br />

⎪⎨ L k + k+1 if 0 ≤ L k + k+1 ≤ s<br />

L k+1 = 0ifL k + k+1 =−1<br />

⎪⎩<br />

s if L k + k+1 >s<br />

Let us denote as F k · the distribution function <strong>of</strong> L k , that is<br />

F k l = PrL k ≤ l l = 0 1s<br />

Furthermore, let us denote as p · the common probability mass function <strong>of</strong> the k s. We<br />

then have<br />

F k+1 l =<br />

l∑<br />

F k l − yp y<br />

y=−1<br />

with F k+1 s = 1. The stationary distribution F · is then obtained as<br />

F l = lim F kl =<br />

k→+<br />

l∑<br />

F l − yp y<br />

with F s = 1. Obviously, the l s are then recovered from 0 = F 0 and<br />

y=−1<br />

l = F l − F l − 1 for l = 1s<br />

The values <strong>of</strong> F l can be computed recursively from the following algorithm:

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

Saved successfully!

Ooh no, something went wrong!