In this paper, we mainly discuss the GARCH models basing on ARCH models in economics.The main work of this article is as follows:First, we briefly review the empirical likelihood estimation ofβ-ARCH model and GARCH model. Then, we get the large sample properties.Second, we discussed the forecast question for ARCH model. Taking MA and ARCH model as an example, the question can be transformed into solving a condition expected expectation value by Bootstrap scheme instead of linear prediction. The results show that the computation using Bootstrap method is simple and accurate.Finally, we introduce the related theory about VaR. Then we integrate the Shanghai and Shenzhen stock composite prices index and give some analysis.Through comparison, we find GARCH(1,1)-GED model is better than others.
|