A Study On The Influence Of Social Media Sentiment On Stock Market

With the increasing development of the Internet,the form and means of media communication have also undergone a series of changes.As an increasingly influential communication medium,social media,represented by microblogging,has gradually evolved into an interactive information sharing and exchange carrier in social activities.Its role as the main information acquisition and communication platform for a large portion of investors can reflect investor sentiment on the one hand and guide investor sentiment on the other.Therefore,the study of the linkage mechanism between social media sentiment and stock market is conducive to clarifying the mutual influence of social media information and stock market,providing a new perspective for China to comprehensively deepen the guidelines of financial market reform,facilitating the state to better regulate the disclosure and dissemination of information in the financial market through tools such as social media,and providing quality soil for the healthy development of China's stock market,thus enhancing the effectiveness of the stock market and advance the securities industry toward the deeper waters of reform.The research in this paper is carried out in several steps: first,by analyzing the ranking of the most influential financial vloggers set by the official Weibo in previous years and the ranking data of third-party evaluation platforms such as watermelon data,37 Weibo financial vloggers with continuous Weibo data from January 2013 to December 2019 are selected.a crawler system is designed by Python language to collect the above-mentioned Weibo vloggers from January2013 to A total of 938,323 microblog messages were collected from the abovementioned microbloggers during the period from January 2013 to December 2019.Then,the textual sentiment analysis method based on sentiment dictionary was used to quantify the crawled microblog text data and construct quantitative sentiment indicators.After that,the 37 time series of quantitative sentiment indicators were then K-Means clustered into 8 groups with low correlation to generate 8 groups of time series quantitative indicators.Then the ADF test and Granger causality test of the data are done to verify whether there is Granger causality between the 8 groups of sentiment indicators and the return of Shanghai Stock Composite Index.Then,the theories and phenomena are sorted out from three aspects of this paper,namely,microblogging sentiment and stock market,above-expected returns and future stock market trend,and stock market sentiment and the accuracy of microblogging sentiment response,and 13 hypotheses are proposed.Finally,model setting and hypothesis testing are conducted for all hypotheses.Through the analysis of the empirical results,this paper mainly obtains the following conclusions:(1)Microblog sentiment refers to indicators that have significant effects on the return,turnover,volatility and next-day return of the Shanghai Stock Composite Index on the same day.For return,there are four types of sentiment indicators showing positive effects at 0.01 significance level.For bloggers with more accurate expressions of sentiment,their positive sentiment blog posts can contribute more to stock market trading behavior.For bloggers whose blog posts are more reflective of the market situation,their sentiment blogging sentiment indicator can negatively affect the volume.For volatility,the more accurate financial bloggers with positive sentiment tendency will have a positive impact on single-day volatility,while for other groups of Weibo financial bloggers,it will have a negative impact.At the same time,the sentiment index of some groups of Weibo financial vloggers can have a positive impact on the next day's return of the Shanghai Stock Exchange Index.(2)If the return of the Shanghai Composite Index exceeds(is lower than)the return reflected by the microblogging sentiment,then this unanticipated increase(decrease)will cause more intense discussion,and to some extent this unanticipated increase(decrease)itself represents a positive(negative)sentiment,which will have a positive(negative)impact on the return of the Shanghai Composite Index the next day.(3)When there is a short-term surge in the market,stock market participants will be more actively involved in the discussion of the stock market,stock market participants spend more time collecting information and disseminating information,making the dissemination of market information more effective,increasing the effectiveness of the market to a certain extent,microblogging quantitative sentiment indicators on the stock market will be more accurate,while the impact on the stock market will also be more stronger.At the same time,long-term rallies will weaken this impact because stock market participants have limited time and energy,and it is difficult to devote themselves to stock market discussions with high intensity for a long period of time.Therefore,when the stock market is in a short-term up state,Twitter sentiment is more accurate in responding to stock market returns,volume,volatility and next-day returns,and when the stock market is in a long-term down state,Twitter sentiment is more accurate in responding to stock market returns,volume,volatility and next-day returns.