Font Size: a A A

Text Sentiment Trend Of Financial News In Listed Companies,Accounting Surplus And Market Reaction:An Empirical Study Based On Shenzhen And Shanghai A-Stock Market

Posted on:2019-03-02Degree:MasterType:Thesis
Country:ChinaCandidate:X ZhaFull Text:PDF
GTID:2348330545477364Subject:Accounting
Abstract/Summary:PDF Full Text Request
The market reaction of accounting surplus has always been a very important topic in the field of empirical accounting research.Existing domestic and foreign related literature mainly focuses on the relationship between accounting surplus and stock excess return.One of its main assumptions is that the accounting earnings information disclosed in the current financial report can change investors'expectations of the future value of the stock and affect the stock's future earnings.However,investors are expected to be affected by other internal and external related text information in addition to the information on the accounting earnings in the financial report.In related external information,financial reports made by major financial media on the company can not only directly affect investors' expectations of the company,but also are important channels for investors to obtain information on accounting earnings information.This article mainly discusses whether the text sentiment tendency of the company's financial news can regulate investors'expectations of accounting earnings digital information.The research on the impact of media reports on the securities market is rapidly developing.It mainly includes three aspects:media influence on investor sentiment,investor sentiment and market reaction,media reports and market reaction.In terms of simplification,it is a transmission mechanism of "media coverasge? investor sentiment ? investor behavior? stock price".In recent years,many researches have been conducted on the mining of financial related texts using text information mining techniques.It has been found that financial related text information has a lot to do with the operation of the securities market.The results of those studies show that financial-related texts,including financial statements,expert speeches,stockholders' remarks,and policy documents of listed companies,all contain potential and valuable information.This information will affect the fluctuation of stock prices,trading volume,and excess returns.It can also reflect the future performance of listed companies.This text utilizes the text mining technology that develops more rapidly in the recent ten years.Firstly,I crawl the financial news of listed companies reported by major media,and use supervised machine learning methods to analyze the emotions of potential text information in financial news.Finally,the annual accumulated text sentiment of financial news of Shenzhen and Shanghai A Share Listed Companies.Based on the market's reaction to the accounting earnings of listed companies,the influence of the potential sentiment of listed companies on the expression of financial news in the market is discussed.The article takes Shenzhen and Shanghai A-share listed companies as samples.Through theoretical and empirical analysis,the study finds that:(1)There is a significant positive correlation between the annual accumulated text sentiment of financial news in Shenzhen and Shanghai A-share listed companies and the excess returns of stocks;(2)The accumulated text sentiment of listed companies' financial news can adjust the correlation between the unanticipated accounting earnings of listed companies and the excess returns of stocks.At the same time,further understanding of the existing comprehensive theories of psychology,finance and other disciplines found that people have different degrees of recognition and response to positive information and negative information,and many literatures consider the impact of negative information on people is greater than positive information;there are also large differences in the stock market performance of listed companies with different media attention.Through the classification analysis and regression of listed companies with positive sentiment and negative sentiment of financial news and listed companies with different news concerns,we studied the adjustment of financial news sentiment under different classification conditions to the market reaction of unanticipated accounting earnings of listed companies is different.The empirical results are consistent with the assumptions:Compared with the listed companies with positive cumulative financial annual sentiment tendency,sentiment tendencies of listed companies with negative sentiment tend to have greater adjustments to correlations between unanticipated earnings of listed companies and excess return on stocks.Compared with the media companies that are concerned about the popularity of the media,the sentiment tendencies of companies with high interest in the media are more reliant on the correlation between the unanticipated earnings of listed companies and the excess return on stocks.Based on the market's response to accounting earnings,the paper explores the influence of potential emotional tendencies on the market response of listed companies' related external information language expressions from a brand new perspective.It considers some external text emotions before the occurrence of a certain event in the future.Whether the accumulation of tendencies will have an impact on the market reaction.The predecessors either individually studied the market reaction of the text sentiment tendencies of the internal and external information of listed companies,or considered the market reaction of accounting surplus separately.Due to frequent financial frauds,investors frequently consider the external information for the forensic role of these data when facing financial data disclosed by the company.Therefore,the sentiment of these external information can adjust the degree of market response to accounting earnings.Because this paper is based on a new perspective,the sentiment tendencies of the research report referred to in this article are not reported at a certain point,but within a year,ten financial media with greater influence exerted news reports on the listed company at different times.There is no authoritative reference for the distribution of weights for the time being.This article can only do a preliminary exploration in this respect.Give the same weight to all financial news throughout the year.Consider from another aspect,this article only considers the data of the corresponding accounting period for a full financial year,because the financial statements report the different dates of the listed companies,and the overall sentiment tendencies of different time periods have not been considered a better measurement method,so the assets have not been transferred.The financial news reported from the balance sheet date to the date of the financial statement is taken into consideration.In the future research,this can be a relatively good measure and a further study can be conducted.
Keywords/Search Tags:Financial news, Text sentiment analysis, Accounting surplus, Market reaction
PDF Full Text Request
Related items