The hierarchical generalized linear model and the bootstrap estimator of the error of prediction of loss reserves in a non-life insurance company

TL;DR

This paper introduces a hierarchical generalized linear model for loss reserving in non-life insurance and proposes a bootstrap estimator to assess prediction error, providing detailed error quantiles via bootstrap methods.

Contribution

It develops a novel bootstrap estimator for prediction error in HGLM-based loss reserving, enabling comprehensive error analysis with practical implementation in R.

Findings

Bootstrap estimator effectively quantifies prediction error.

Quantiles of absolute prediction error provide detailed error insights.

Method enhances accuracy of loss reserve predictions.

Abstract

This paper presents the hierarchical generalized linear model (HGLM) for loss reserving in a non-life insurance company. Because in this case the error of prediction is expressed by a complex analytical formula, the error bootstrap estimator is proposed instead. Moreover, the bootstrap procedure is used to obtain full information about the error by applying quantiles of the absolute prediction error. The full R code is available on the Github https://github.com/woali/BootErrorLossReserveHGLM.