Decomposition Of Stock Market Volatility Based On Investor Sentiment And Macroeconomic Fundamentals

Chinese stock market has been dubbed "policy market" and "retail city" title,this statement has its rationality to a certain extent,the investors in the majority with retail market in our country,the majority of people accept the limited investment in education,but also cannot be denied that the stock market fluctuations in the very great degree reflects the fact that the macroeconomic operation.In fact,with the deepening of the research on the stock market,behavioral finance thinks investors emotions influence their trading behavior,and then affect market movements.And traditional financial theory is that share price is around its intrinsic value fluctuates up and down the claim still has its rationality.At the same time,more and more people realize that investor sentiment and macroeconomic fluctuations might be to promote the main force behind the stock market volatility.Combined long and short-term fluctuations,however,considering the lack of proper empirical method.Based on the above consideration,this article adopts the model of GARCH-MIDAS attracted much attention in recent years as an empirical framework of this paper,considering the mechanism of action of stock market fluctuations of both short-term and long-term two levels.Also used in the process of determination of investor sentiment proxy variables of stock market turnover rate used by the academic circles as a benchmark,at the same time,through combination of web crawler and text mining,the author attempts to improve investor sentiment in the short term factors on the contribution of the model,the integrated imperial Python,R language programming to realize network text of the text of the information collection,processing,cleaning,word segmentation research steps,such as investor sentiment index constructed directly from the network text.Through a series of theoretical and empirical studies,this paper get the following conclusions: first,the long and short-term volatility is divided into two components on this knowledge,the performance of the domestic and foreign stock market volatility is consistent: the stock market volatility include composition of long-term and short-term into two levels,and macroeconomic information contained in the stock market volatility composition of driving force for a long time.At the same time,the composition of the short-term volatility of the stock market in China can get very good explanation by investor sentiment.Second,through a combination of web crawler and text mining to build sentiment index is good,for follow-up study of index build greatly broaden the channels of information collection.Third,the emotional dictionary in the building of sentiment index has not mentioned the important role.Emotional dictionary word choice more comprehensive,screening more targeted,eventually build investor sentiment index is more representative.Especially involves the professional in the field of sentiment analysis,need a rich corpus of the area as well as the method of establishing scientific and reasonable sentiment index.Fourth,the GARCH model and MIDAS(Mixed Frequency Data Sampling)GARCH model and construct-MIDAS model can effectively expand the research in depth,at the same time can also be mixed frequency according to the maximum extent,mining information,and its application in our country stock market volatility composition model research is effective and applicable.