A MIXTURE MODEL FOR PAYMENTS AND PAYMENT NUMBERS IN CLAIMS RESERVING. Issue 1 (17th October 2016)

• Record Type: Journal Article
• Authors: Gigante, Patrizia; Picech, Liviana; Sigalotti, Luciano
• Is Part Of: ASTIN bulletin. Volume 48:Issue 1(2018)
• Page Start: 25
• Page End: 53
• Publication Date: 2016-10-17
• Subjects: Claims reserving, -- Tweedie's compound Poisson model, -- conditional mean square error of prediction, -- hierarchical generalized linear models, -- h-likelihood; Insurance -- Mathematics -- Periodicals; Risk (Insurance) -- Mathematics -- Periodicals; 368.01
• DOI: 10.1017/asb.2016.30
• Languages: English
• ISSNs: 0515-0361

Abstract

Abstract: We consider a Tweedie's compound Poisson regression model with fixed and random effects, to describe the payment numbers and the incremental payments, jointly, in claims reserving. The parameter estimates are obtained within the framework of hierarchical generalized linear models, by applying the h -likelihood approach. Regression structures are allowed for the means and also for the dispersions. Predictions and prediction errors of the claims reserves are evaluated. Through the parameters of the distributions of the random effects, some external information (e.g. a development pattern of industry wide-data) can be incorporated into the model. A numerical example shows the impact of external data on the reserve and prediction error evaluations.