this code is used to construct a 2T-PoT-MIDAS-Copula model to predict volatility marginally and VaR in joint
Zhenyu Yang, School of Statistics and Mathematics, Zhejiang Gongshang University, China
We proposes a dynamic two-tailed peak over threshold (2T-PoT) model based on the dynamic bivariate peak over threshold model. The proposed model has a double-tailed distribution with Pareto behavior. We combine the extended model with the GARCH-MIDAS model for out-of-sample volatility forecasting. The empirical results show that the extended model has better volatility forecasting ability than the original dynamic PoT model. Further, we define a 2PoT-GAS-Copula model similar to PoT-GAS-Copula, and the performance of the former in predicting VaR is consistent with the latter, illustrating the reliability of the new model.
If you want to use the code, comment the source: https://github.com/yuyuliangyu/2T-PoT-MIDAS-Copula