Dynamic risk measurement of futures based on wavelet theory

In this paper, we choose consecutive month contract of natural rubber in China futures market for the study. Based on the GARCH class model, we combine the wavelet analysis with extreme value theory to get the approximate distribution of time series and then use rolling time window to predict dynamic value at risk. The empirical results show that the models all have good predictive ability, and the models which using wavelet analysis to estimate the threshold in generalized Pareto distribution achieve a better dynamic prediction.