No.636 Robust estimation of a joint risk regression model with application to earthquake catastrophe risk analysis

本期主题

Robust estimation of a joint risk regression model with application to earthquake catastrophe risk analysis

Value at Risk (VaR) and Conditional Value at Risk (CvaR) or Conditional Tail Expectation (CTE) are important risk measures in financial and insurance risk analysis. In this paper, a generalized form of joint VaR and CTE regressions is proposed under a joint scoring function. A robust estimation method for the proposed model is discussed. The estimation procedure consists of two steps: the first step considers the estimation of the VaR model and the second step focuses on the estimation of the CTE regression. In order to obtain a robust estimate, the Huber loss is applied in the second step. To illustrate the application of the proposed model, the earthquake catastrophe risk is analyzed in the empirical study, the results of which can be further used to design effective risk-sharing mechanisms.