CFE 2015: Start Registration
View Submission - CFE
A1371
Topic: Contributions on time series Title: A copula quantile approach to conditional Value-at-Risk estimation Authors:  Krenar Avdulaj - Department of Econometrics-IITA-The Czech Academy of Sciences (Czech Republic) [presenting]
Jozef Barunik - UTIA AV CR vvi (Czech Republic)
Abstract: We propose to use copula quantile regression models and realized volatility to estimate Value-at-Risk (VaR) of an institution conditional on some other institution being under financial distress. The proposed model uses copulas from elliptical family, Normal and $t$ copula, and the realized volatility measure which is calculated using 5 minutes returns. Contrary to the literature which studies the systemic risk of financial institutions, we estimate the risk contribution of an institution to some other institution. We apply the model on 21 most liquid U.S. stocks from seven main market sectors. We find that stocks from financial sector have the highest risk transmission among each other, followed by the Information Technology sector. Consumer Staples industry stocks and Health Care have the lowest risk transmission.