Special Data Science Seminar - Seul Ki Kang - Risk Analysis and Uncertainty Quantification in Insurance Ratemaking
Title: Risk Analysis and Uncertainty Quantification in Insurance Ratemaking
Abstract: Consideration of the modern general risk measures beyond the variance into premium calculation is attracting much attention in the insurance industry and actuarial literatures. Especially many recent literatures documented a premium principal based on a quantile risk measure which has advantages of more accurate and better measurement of the variability of a risk class, and such advantages are essentially important for insurance industry under Solvency II Regulation System. Following the stream of such studies, in our research, we are interested in forecasting two widely employed risk measures of claim losses, which are the Value-at-Risk (VaR) and Expected Shortfall (ES), conditional on some characteristics of policies. We propose several novel ways of developing efficient statistical methods for forecasting such risk measures and making inference, which can be applied to insurance ratemaking data. Moreover, we suggest innovative methods of quantifying uncertainty of our proposed inference methods. Data analysis using real insurance data show that our proposed methods are quite efficient. Additionally, a simulation study confirms a good finite sample performance of the proposed methods.
Meeting ID: 949 5373 0428