Analysis Of Chinese GDP Based On ARIMAX's Intervention And Co-integrated Model

The outbreak of the COVID-19 in 2020 has brought huge losses to our country's economic development.But due to insufficient data,it is impossible to use the existing time series model to explore the specific impact of the COVID-19.Therefore,in response to the 2008 financial crisis,this paper first selects the explanatory variable that has the greatest impact on the current quarterly GDP data through the lasso method:the added value of the tertiary industry;then,an intervention model based on the ARIMAX model is established for our country's current quarterly GDP data from 1993 to2019.This intervention model has a better fitting effect than the traditional intervention model based on the ARIMA model,thereby better portraying the impact of the financial crisis on our country's current quarterly GDP.In 2021,the global epidemic has been brought under certain control,and our country's economy has also fully recovered.Therefore,this paper establishes an ARIMA(1,1,2)model for our country's GDP data from 1979 to 2017 through the steps of white noise and stationarity test,preliminary identification of the model,specific model order,model estimation,and model test.Then using the lasso method to select 15 variables in the lagging period,excluding the multi-collinearity between the variables,and finally select three indicators that have the greatest impact on GDP: general public budget income,per capita disposable income of urban residents and industrial growth value.The three selected variables lagging one period are used as input variables to establish a co-integration model for GDP.Through test analysis and comparison in 2018 and 2019,the co-integration model with lagging one-period explanatory variables has a better effect on short-term prediction.Finally,the co-integration model is used to predict our country's GDP in 2021.