Bayesian multivariate Poisson models for insurance ratemaking

When actuaries face the problem of pricing an insurance contract that contains different types of coverage, such as a motor insurance or a homeowner's insurance policy, they usually assume that types of claim are independent. However, this assumption may not be realistic: several studies have shown that there is a positive correlation between types of claim. Here we introduce different multivariate Poisson regression models in order to relax the independence assumption, including zero-inflated models to account for excess of zeros and overdispersion. These models have been largely ignored to date, mainly because of their computational difficulties. Bayesian inference based on MCMC helps to resolve this problem (and also allows us to derive, for several quantities of interest, posterior summaries to account for uncertainty). Finally, these models are applied to an automobile insurance claims database with three different types of claim. We analyse the consequences for pure and loaded premiums when the independence assumption is relaxed by using different multivariate Poisson regression models together with their zero-inflated versions.

[1]  Efthymios G. Tsionas,et al.  BAYESIAN MULTIVARIATE POISSON REGRESSION , 2001 .

[2]  Bradley P. Carlin,et al.  Bayesian measures of model complexity and fit , 2002 .

[3]  Peng Shi,et al.  Actuarial Applications of a Hierarchical Insurance Claims Model , 2009, ASTIN Bulletin.

[4]  Michel Denuit,et al.  Number of Accidents or Number of Claims? An Approach with Zero-Inflated Poisson Models for Panel Data , 2009 .

[5]  Dankmar Böhning,et al.  The zero‐inflated Poisson model and the decayed, missing and filled teeth index in dental epidemiology , 1999 .

[6]  D. Matem.  A priori ratemaking using bivariate Poisson regression models , 2008 .

[7]  Montserrat Guillén,et al.  Bonus-Malus Scales in Segmented Tariffs With Stochastic Migration Between Segments , 2003 .

[8]  Hans van Ophem,et al.  A GENERAL METHOD TO ESTIMATE CORRELATED DISCRETE RANDOM VARIABLES , 1999, Econometric Theory.

[9]  Montserrat Guillén,et al.  On the link between credibility and frequency premium , 2008 .

[10]  Montserrat Guillén,et al.  Long-range contagion in automobile insurance data: estimation and implications for experience rating , 2000 .

[11]  Multivariate Poisson Regression with Full Covariance Structure , .

[12]  Loukia Meligkotsidou,et al.  Multivariate Poisson regression with covariance structure , 2005, Stat. Comput..

[13]  Michel Denuit,et al.  Credibility premiums for the zero-inflated Poisson model and new hunger for bonus interpretation , 2008 .

[14]  Robert Kohn,et al.  Multivariate probit models for conditional claim-types , 2009 .

[15]  John M. Olin.  Markov Chain Monte Carlo Analysis of Correlated Count Data , 2003 .

[16]  D. Karlis,et al.  Finite mixtures of multivariate Poisson distributions with application , 2007 .

[17]  Montserrat Guillén,et al.  Allowance for the Age of Claims in Bonus-Malus Systems* , 2001, ASTIN Bulletin.

[18]  Dimitris Karlis,et al.  Finite normal mixture copulas for multivariate discrete data modeling , 2009 .

[19]  Michel Denuit,et al.  Fixed versus Random Effects in Poisson Regression Models for Claim Counts: A Case Study with Motor Insurance , 2006, ASTIN Bulletin.

[20]  Pravin K. Trivedi,et al.  Regression Analysis of Count Data: Preface , 1998 .

[21]  D. Karlis,et al.  Analysis of sports data by using bivariate Poisson models , 2003 .

[22]  Montserrat Guillén,et al.  Time-varying credibility for frequency risk models: estimation and tests for autoregressive specifications on the random effects , 2003 .

[23]  Michel Denuit,et al.  Actuarial Modelling of Claim Counts , 2007 .

[24]  Michel Denuit,et al.  Risk Classification for Claim Counts , 2007 .

[25]  Jye-Chyi Lu,et al.  Multivariate zero-inflated Poisson models and their applications , 1999 .

[26]  P. Berkhout,et al.  A bivariate Poisson count data model using conditional probabilities , 2004 .

[27]  Emiliano A. Valdez,et al.  Hierarchical Insurance Claims Modeling , 2008 .

[28]  N. L. Johnson,et al.  Discrete Multivariate Distributions , 1998 .

[29]  Hilippe,et al.  RISK CLASSIFICATION FOR CLAIM COUNTS : A COMPARATIVE ANALYSIS OF VARIOUS ZERO-INFLATED MIXED POISSON AND HURDLE MODELS , 2006 .

[30]  D. Karlis,et al.  Bivariate Poisson and Diagonal Inflated Bivariate Poisson Regression Models in R , 2005 .

[31]  C. Dean.  Testing for Overdispersion in Poisson and Binomial Regression Models , 1992 .

[32]  Kui Wang,et al.  A bivariate zero-inflated Poisson regression model to analyze occupational injuries. , 2003, Accident; analysis and prevention.