Use this URL to cite or link to this record in EThOS: http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.650812
Title: A study of actuarial models for insurance based applications
Author: Butt, Zoltan
ISNI:       0000 0004 5357 6628
Awarding Body: City, University of London
Current Institution: City, University of London
Date of Award: 2014
Availability of Full Text:
Access from EThOS:
Access from Institution:
Abstract:
Actuarial aspects of two important fields of insurance are considered: calculating personal injury damages for working age adults (liability insurance) and measuring the mortality in insurance based populations (e.g. life insurance and pensions). The contribution of the thesis is to demonstrate a wide array of modelling techniques and their practical implementation in these two key areas of actuarial science. The first part considers the modelling of the labour force dynamics from the perspective of the loss of earnings multipliers in England and Wales. It reviews the estimation methods of involuntary non{participation in the labour market in relation to future loss of earnings. In response, a robust multiple state modelling methodology is developed that allows conditioning on personal characteristics of working age individuals such as disability, educational attainment and the current employment state. Applied to UK longitudinal Labour Force Survey data, it quantifies the disadvantages that plaintiffs with post-injury earnings capacity face in the labour market. This practical modelling framework leads to a set of improved loss of earnings multipliers in subsequent editions of the Ogden Tables now used in the Courts in England and Wales. The second part focuses on the modelling and estimation of mortality rates using Poisson likelihood maximisation methods. In terms of graduation, it undertakes a comprehensive assessment of the frailty models and their implications. Then it puts forward suitable parametric modelling structures in order to measure the scale of individual heterogeneity and applies generalised linear modelling graduation techniques to a large array of insurance based mortality data. In terms of forecasting, it considers the generalised Lee-Carter type modelling structures of Renshaw and Haberman(2006) and demonstrates their suitability for practical applications. Furthermore, it develops a novel stratified Lee-Carter model for the measurement of the effects of explanatory factors (other than age and time). An efficient programming package in R is provided for this class of modelling framework. Finally, a detailed analysis of the mortality trends observed in private pension scheme data serves as a case study.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID: uk.bl.ethos.650812  DOI: Not available
Keywords: HB Economic Theory
Share: