Abstract
Is an actuarial rating of bonus-malus type based on the parametric credibility model of mixed Poisson distributions too differentiated? To answer this question, we enlarge the model of mixed Poisson distributions by considering mixed negative binomial distributions. The correlation coefficient between the annual claim numbers for an individual can then be adjusted, and we calculate the differences between the ratings coming from this new model of parametric credibility. An application of these results to a car portfolio shows that adjustment of the correlation may yield significantly smaller rating differences than those coming from the classical model. To cite this article: D. Pierre-Loti-Viaud, C. R. Acad. Sci. Paris, Ser. I 339 (2004).
Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
