Application Research Based On Extended Intervention Analysis Model

Based on the VAR model and the extended intervention model,this paper studies the macroeconomic indicators CPI and M2 from 1991 to 2020,and the impact of intervention policies affecting small and medium investors on the quarterly data of the Shanghai Composite Index,and establishes six different generalized intervention models.Finally,the optimal intervention model is determined by comparison,and the results show that the intervention model promoted from the perspective of macro economy is the best.And it is applied to analyze the effect of intervention policies affecting small and medium investors on Shanghai Stock Index in each stage.It enables investors to study and judge the general trend to grasp the direction,and provides a scientific decision-making basis for rational investment.The details are as follows:1.The explanatory variables CPI and M2 and the explained variable SSE index SAZA were selected.Based on the research of scholars,the macro-economic data under the new normal during 1991-2004 and 2013 were added to establish whether there were new changes in the relationship between VAR and impulse response description.The empirical results show that in the long run,the quarterly change of CPI has a negative transmission effect on SAZA,while the quarterly change rate of M2 has a positive transmission effect on SAZA.The variance decomposition shows that the impact contribution rate of CPI is greater,and the impact and duration are basically the same.The existence of the correlation between the explained variables and the explained variables is proved,which is the basis for the subsequent establishment of the extension model.2.After considering the influence of prediction accuracy and other reasons,three stages were divided into three stages based on the previous scholars' periodic evolution,AREGARCH-M and other multi-angle studies.Under the background of the second and third stages of macroeconomic development,the intervention model of the period most affected by intervention was established.The main trend models are based on ordinary regression and ADLM respectively,and there are six models in total.(1)Through the comparison of multiple prediction error indicators,the prediction and performance of the extended intervention analysis model based on ADLM with macroeconomic variables were the best in each stage of division.That is to say,the prediction of the intervention extension model from the macroeconomic perspective in this paper is better.(2)The six intervention models all proved the results of impulse response analysis.(3)Representative intervention policy analysis: At the beginning of 2005,the intervention policy of non-tradable shares reform had a slightly stronger effect on the Shanghai Composite Index and the effect lasted for a long time;The impact of the "New Nine Policies" intervention policies in 2014 was shorter and the intensity was slightly lower than that in 2005.Based on the results and nature of these two intervention policies,the future favorable policies favoring medium and small investors will drive the stock market to rise.