Research On The Forecast Of SSE Composite Index Return Based On Micro Index And Baidu Index

Internet data records investors' micro-emotional information and search attention,it also provides a large amount of data base and new research perspective to study the macro operation rules of the stock market.In recent years,research based on Internet data to explore investor sentiment,attention and stock market relations has attracted more and more attention from domestic and foreign scholars.Based on existing research,this paper respectively constructs a bullish index on investor sentiment and attention in the stock market by using Sina microblog and Baidu search engine of two different types' platform data.From the perspective of information supply and demand,combining with the knowledge of informatics and behavioral finance,this paper explained the different meanings of the two representatives,and discussed the time lag relationship between them and Shanghai composite index return.Through theoretical analysis and empirical test,this paper found that social platforms such as microblog and network search engines are the carrier of information,people express their views and emotions on social networking sites such as microblog,which reflect the information supply behavior,reflecting investor sentiment.While people search for information on network search engines is information demand behavior,reflecting the attention of investors.Both of them affect the trading behavior of the stock market through the bridge of "information supply and demand".Weibo bullish index built by micro index is ahead of Baidu's bullish index(Market attention).In addition,the Weibo bullish index has a predictive effect on the Shanghai composite index return.Therefore,the micro-index platform can be used as an effective supplement to predict stock market return.The Baidu bullish index shows a significant positive correlation with the stock market yield,but it has little effect on the yield forecast.Further,this paper also use time-varying parameter vector autoregressive(TVP-VAR)model of interval impulse response function and timing pulse response function to capture the time-varying characteristics between investor sentiment,attention and stock market return.The empirical analysis results show that the influence of investor sentiment and concern on stock market returns has obvious time-varying characteristics:(1)Investor sentiment has positive effect on stock market return,and the short-term effect is more obvious.(2)Investor attention has little effect on stock market returns;In turn,the impact of stock market return on investors' attention is more significant and stable.(3)When the stock market fluctuates abnormally,the impact of investor sentiment on stock market returns will be more pronounced than when the stock market is stable,that is,investor sentiment has contributed to the stock market panic decline or irrational prosperity.