GLUE-VALUE-AT-RISK ON AGGREGATE LOSS SEVERITY DISTRIBUTION

• Main Author: Akbar Alfan, Raihan
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/78417
• Institution: Institut Teknologi Bandung

A crucial tool in making future large loss predictions is the measure of risk. In real life, Value-at-Risk (VaR) and Expected Shortfall (ES) are used as risk measurements to compute losses. This study will introduce GlueVaR, a risk measure that is a linear combination of VaR and ES. The case study of the aggregate loss severity of motor vehicle accident insurance, which consists of property insurance claims and medical expenses insurance claims, will be used to calculate the size of this risk. Three different approaches or methods, namely the Non-Parametric method, Non-Parametric method with Monte Carlo, and Parametric with Monte Carlo, were used to perform the calculations. Evaluation of the risk measurement prediction results is carried out by calculating the coverage opportunity and MSEP of each predictor. The results show that the estimation results produced more accurate results and smaller error values by using the Monte Carlo simulation on the Estimative VaR