Modelling For Heteroskedasticity

Since Engle propose ARCH model to describe the volatility in time series in1982. a large number of researchers are improving and extending the model on the foundation. Other than directly modelling the conditional heteroskedasticity before, we propose a new model for the inverse of the conditional heteroskedasticity in this paper. MLE is used to estimate the parameters of the new model, then we also give a simple,effective and quick computational algorithm. At last,we try to use the proposed model to analyse the quarter GDP data of China and the USA.