Template-type: ReDIF-Paper 1.0
Author-Name: J. Harnau
Author-Workplace-Name: Department of Economics and Oriel College, University of Oxford
Author-Email: jonas.harnau@oriel.ox.ac.uk
Author-Name: B. Nielsen
Author-Workplace-Name: Department of Economics and Nuffield, University of Oxford
Author-Email: bent.nielsen@nuffield.ox.ac.uk
Title: Over-dispersed age-period-cohort models
Abstract: We consider inference and forecasting for aggregate data organised in a twoway table with age and cohort as indices, but without measures of exposure. This is
modelled using a Poisson likelihood with an age-period-cohort structure for the mean while allowing for over-dispersion. We propose a repetitive structure that keeps the
dimension of table fixed while increasing the latent exposure. For this we use a class of infinitely divisible distributions which include a variety of compound Poisson models
and Poisson mixture models. This results in asymptotic F inference and t forecast distributions 
Keywords: Chain-ladder model; Forecasting; Generalized linear model; Inference; Infinitely divisible, Two-way table
Length: 22 pages
Creation-Date: 2017-07-31
Number: 2017-W06
File-URL: https://www.nuffield.ox.ac.uk/economics/Papers/2017/HarnauNielsen2017apcDP.pdf
File-Format: application/pdf
Handle: RePEc:nuf:econwp:1706