Study On Estimation Method Of Outstanding Claims Reserve Based On Generalized Linear Mixed Models

As an important item of debts, the estimation of the level of outstanding Claims Reserve impacts on multiple aspects of the operation of the insurance company:dividend declaration, solvency, insurance product pricing and tax payments. So the estimate of outstanding Claims Reserve have been paid more attention by Non-insurance company and regulatory authorities.In last two decades, stochastic models for the estimation of outstanding Claims Reserve have been a hotspot. As the most widely applied stochastic models, generalized linear models (GLMs) have been extended the distributions of response variable from normal to the exponential family, based on the classical linear models. As a result of their adaptive characteristics, Generalized linear models have received a lot of attention in non-life actuary. GLMs require a sample of independent random variables. However, in many actuarial and general statistical problems, the assumption of independence is not full filled. Such as longitudinal, binary, spatial or clustered data.This article puts focus on longitudinal data which are repeated measurements on a group of subjects over time. Since they share subject-specific characteristics, observations on the same subject over time often are substantively correlated and require an appropriate toolbox for statistical modeling. Generalized linear mixed models (GLMMs), based on GLMs, include random effects in the linear predictor, which response the correlation between observations on the same subject, and the heterogeneity among subjects.This paper studies parameter estimation, inference and prediction of GLMMs, and discusses applications of GLMMs in the estimation of outstanding Claims Reserve.