Prediction of Outstanding Claim Using Stochastics Chain Ladder Method

• Main Author: FADLI AZIM (NIM:20811002); Pembimbing : Dumaria R. Tampubolon, Ph.D, MUHAMAD
• Format: Theses
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/20303
• Institution: Institut Teknologi Bandung

A runoff triangle may be used to organize data from a long-tail insurance business based on accident and development years. The chain ladder method is often used to predict the outstanding claims of such data. One prediction problem is the consistency of the prediction. In this thesis, bootstrap is used to create pseudo triangles. The outstanding claims of each of those pseudo triangles and its corresponding prediction error are calculated. The mean of the the outstanding claims predictions as well as the mean of the prediction errors of the pseudo triangle are then compared with those of the original runoff triangle. <br /> <br /> <br /> <br /> The data used in this thesis is from a property insurance business of a general insurance company, called company ABC. The incremental claim amount data is assumed to follow an Over-Dispersed (ODP) distribution, with a natural logarithm link function. It is obtained that the prediction of the outstanding claims is approximately 15.190 billion rupiah with its prediction error of approximately 11.969 billion rupiah. Out of 30 pseudo triangles, it is obtained that the mean of the predicted outstanding claims is approximately 15.230 billion rupiah with its mean prediction error of approximately 9.698 billion rupiah. <br /> <br />