THE PREDICTION OF THE REPORTED BUT NOT SETTLED (RBNS) AND THE INCURRED BUT NOT REPORTED (IBNR) CLAIMS RESERVES USING THE DOUBLE CHAIN LADDER (DCL) METHOD

• Main Author: MADONNA (NIM: 20814007), NORA
• Format: Theses
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/23521
• Institution: Institut Teknologi Bandung

An insurance company is required to set up a claims reserve to pay claims reported by the policyholders. In a long tail insurance business, there is a delay between when a claim occurred and when it is settled and the delay may take more than one year. There are two types of outstanding claims: Incurred But Not Reported (IBNR) claims; and Reported But Not Settled (RBNS) claims. The Chain Ladder Method (CLM) is often used to estimate the outstanding claims reserve for a long tail insurance business. The data in a long tail insurance business may be represented in a runoff triangle format. In CLM, the estimate of the outstanding claims reserve may be determined using the runoff triangle of paid claims by including all sources of delay into a development pattern and it will only calculated a predictive value of the claims reserve. The method used in this thesis is closely related to CLM and is called a Double Chain Ladder (DCL) method. The DCL method applies the Chain Ladder algorithm to two run-off triangles, namely: the run-off triangle of paid claims; and the run-off triangle of the number of reported claims. The additional information from the run-off triangle of the number of claims reported and the interpretation in the context of insurance data, enable the DCL method to estimate the RBNS and IBNR reserves separately. To see a significant difference between the two methods, and to measure the variability of the claims reserve, the prediction errors given by both methods are determined using bootstrapping techniques. The analysis showed that the prediction error given by the DCL method is less than that given by CLM.