Finance Case Analysis Based On Transmissibility Function Model And GARCH Model

For the case of time series with exogenous variables, it is better to set up transmissibility function model (TFM). TFMs have been widely applied to the fields of economy,finance,industrial production and engineering, but is not so common in the field of stock. In this paper, the TFM is used to fit and forecast Shanghai composite index, which gets satisfying results.In order to compare the fitting and forecasting errors of different models, an average method of many periods'and multiple steps'fitting and forecasting errors is used in this paper. This method may decrease the chance of fitting error and forecasting error so as to increase the credence of the results. Compared with the fitting and forecasting errors of other models, TFMs have better results than traditional ARMA models and GARCH models.For the time series with highly interrelated exogenous variables, we suggest a new method with weighted exogenous variables. The models set up by this method are advantageous than the models with all exogenous variables or the models keeping only one exogenous variables. The case in paper verifies this conclusion.In this paper, a new model—transmissibility function model with conditional heteroscedasticity is proposed, and its conditions as well as the fitting and forecasting effect have been studied. We find this model satisfactory, especially for the time series with significant heteroscedasticity.