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.

<strong>Actuarial</strong> Analysis <strong>of</strong> the French Bonus-Malus System 337<br />

a 25 = 10 %<br />

a 30 = 10 %<br />

a t = 0 for all other t<br />

The minimization <strong>of</strong><br />

∑<br />

E A = a t t <br />

t=1<br />

with respect to and then gives the optimal solutions<br />

= 00710 and = 00133<br />

The financial equilibrium is achieved, as the total premium income tends to 104.29 % <strong>of</strong> the<br />

initial one. When working with a weighted average <strong>of</strong> the t s, the values associated<br />

with large t play the prominent role, resulting in values for and similar to those obtained<br />

for t>20 in Table 9.1.<br />

With optimal CRM coefficients, the discount for claim-free policyholders is rather modest<br />

(1.33 % per claim-free year), but the penalty in case <strong>of</strong> a claim is also moderate (7.1 %). The<br />

large differences compared with the <strong>of</strong>ficial values <strong>of</strong> today’s bonus-malus system in France<br />

(5 % <strong>of</strong> discount per claim-free year, and 25 % increase per claim) can be explained by the<br />

fact that all the penalties are suppressed after two claim-free years according to the terms <strong>of</strong><br />

the French law, which is particularly generous.<br />

We have also tested two different sets <strong>of</strong> a t s, to study the influence <strong>of</strong> the age structure <strong>of</strong><br />

the portfolio on the optimal CRM coefficients. With the age structure <strong>of</strong> an ‘old’ portfolio,<br />

that is,<br />

the minimization <strong>of</strong> E A gives<br />

a 1 = 10 %<br />

a 5 = 10 %<br />

a 12 = 10 %<br />

a 20 = 20 %<br />

a 25 = 20 %<br />

a 30 = 30 %<br />

a t = 0 for all other t<br />

= 00658 and = 00115<br />

With the age structure <strong>of</strong> a ‘young’ portfolio, that is,<br />

a 1 = 30 %<br />

a 5 = 20 %

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

Saved successfully!

Ooh no, something went wrong!