Researching For Modeling And Application On Time Series Of Savings Deposits Of Residents

The purpose of this thesis is to research the savings deposit model of our country. A scientific and rational savings deposit model is built based on time series method.First, the basic theory of time series is discussed. The building, identification, parameter estimating, test and forecasting of transfer function model are analyzed deeply. The problem of the savings deposit models is summarized and proposed.This thesis demonstrates several influencing factors of savings deposit. Then, the most important factor is imported to the transfer function model, and the transfer function model of savings deposit is built combining the statistic software SAS and Eviews. The result shows that the forecasting effect of transfer function model is better than the single variable ARIMA model by Comparing the transfer function model and single variable ARIMA model. It improves the rationality and validity of transfer function model with explaining variable which is the most important influencing factor, and it also shows that transfer function model is more conformable to forecast savings deposit of our country. In the last, the forecasting results are analyzed and some polices and advice are proposed based on the actual economy of our country, which is valuable.