Empirical Analysis Of The Impact Of Internet Financial News On Stock

At the end of twentieth century, Internet-based news media was celebrated as the fourth media following the newspapers, radio and television by the United Nations Information Centers. Until now, Internet-based news often comes at the right time and it is very easy to access, so Internet-based news becomes more and more popular. Especially, Internet-based financial news often tells something about the state of operation and financial condition of a company, which is very important for Chinese investors. And researches have shown that investors need more asset information than assets of their own and investors can obtain more investment profit when they get enough asset information. So, it will be very significant to study the relationship of Internet-based financial news and stock return.This paper manages to study the following two problems. First, how can we mining the mass of unstructured Internet-based financial news in order to find useful information for investors? Second, How about the influence of different kinds of Internet-based financial news on stock?In this paper, text mining technique and financial research method are used to address the two problems. First, we study text classification technique which is an important branch of text mining technique. Second, the Internet-based financial news is classified into three categories, including merger category, profitability category and refunding category by the method of text classification technique. Third, we use event study method to explore the impact of different categories of news on stock return. At last, we make some policy advices according to the study conclusions.This thesis consists of six parts. The first part:introduction. This part introduces the study background and a thorough review of research at both home and abroad. Then we introduce our research contents and methods. The second part:text mining technique. This part introduces the connotation of text mining and illustrates the process of text mining. The third part:the research about text classification of Internet-based financial news. This part introduces the principle of text classification technique, including the key techniques and key classification methods. Then, we classify the Internet-based financial news into three categories by the methods of text classification technique. The fourth part:the measurement of influence of the Internet-based financial news on stock return. This part mainly introduces the principle and process of event study method, which is used to study the relationship of the Internet-based financial news and stock return. The fifth part: the empirical analysis of the impact of internet-based financial news on stock return. This part is very important in the whole paper. First, we introduce our stock data and news sample. Then we analyze the impact of different categories of internet-based financial news on stock return respectively. The last part: conclusion. In this part, first, we conclude what we have done in this paper and what conclusions we have made. Then, we make some policy advices according to the empirical results. Last, we conclude the shortage of this paper and give some suggestions for further study.In this paper, we have some interesting findings.(1) News of merger category and profitability category can have a positive impact on the stock return.(2) Merger news has strong impact on the stock return from the fourth day before the news release to the ninth day after the news release. We guess maybe investors have obtained the information before the news release. Investors can get more benefits if they buy stock before the news release.(3) Profitability news has stronger impact on stock than merger news. Especially, profitability news has strong impact on the stock price on the day of news release.(4) Refunding news has little negative impact on the stock return, which is different from that of merger and profitability news.Base on the study conclusions, we make some suggestions.(1) Because merger news can have a positive impact on the stock return, the government should give some support to the listing companies which will take the merger action.(2) The government should take some measures to keep inside information secrete during the merger and acquisition of listing company.(3) Because investors blindly believes refunding is not good for their benefits, the government should enhance information disclosure about refunding, which can help investors make right decisions.This paper makes some innovations in the research method and research direction. First, we combine information technique and financial research methods, which is research methods innovation. Second, we try to classify Internet financial news into different categories and study the impact of news on stock return, which is research direction innovation.