Abstract:
To make up for the defect of the existing VaR calculation model in the process of calculating the VaR value of multiple exchange rate risk factor, the authors applied the Panel-GARCH model to the measurement of the VaR of exchange rate risk. When the Panel-GARCH model is compared with the univariate GARCH model, BEKK model and DCC model in the multivariate GARCH model, the authors found that the result of simultaneous VaR calculation is better than the other three models. The panel GARCH model based on the assumption of the residual normal distribution can better capture exchange rate fluctuations, and its application can enhance the accuracy of VaR estimate and interest rate risk management of financial institutions or enterprises