Semantic Portrayal Of Financial Media And Its Impact On Financial Markets

Preventing and resolving major risks is one of the three major battles to build a moderately prosperous society in the new era.In today’s era of information bombardment,it is important and crucial to prevent and resolve major risks by tapping into and protecting the stable development of the financial market with high-quality information transmission.With this motive,this paper constructs a "financial media semantic" index based on the content of domestic news text data,and examines the index’s predictive function and intrinsic mechanism in the financial market from three dimensions: overall financial market prosperity,stock returns of sub-markets and individual stock price fluctuations,and analyzes the index’s predictive function and intrinsic mechanism from the perspective of financial system stress.The potential impact of the semantic index of financial media in financial vulnerability is analyzed to test the important role of the index in preventing and resolving major financial risks.The main contents and conclusions of this paper are as follows.First,differing from the support vector machine(SVM)model and the long and short-term memory neural network(LSTM)training model often used in previous studies,this paper adopts a more effective and accurate bidirectional long-term short-term memory neural network(Bi LSTM)model to train news texts to obtain the degree of semantic tendency of financial media,and thus constructs the baseline data of the semantic index of financial media,while using the support SVM model was used to obtain the reference indices for comparison with the prediction effect.The empirical analysis shows that the financial media semantic index constructed in this paper can respond more effectively to the timely fluctuation characteristics of the stock market as a market impact factor than the policy uncertainty index and the composite investor sentiment index.Second,the predictive function and intrinsic mechanism of the financial media semantic index on the financial market are searched from three dimensions:(1)using monthly data of macroeconomic sentiment,securities investor confidence and consumer confidence index as the macro reflection of the financial market sentiment,matching the daily financial media semantic index data,the study finds that(2)Using GARCH-MIDAS and GARCH models to empirically analyze the impact of long-and short-term financial media semantics on the stock returns of the SSE 50,CSI 300 and CSI 500 submarkets,we find that the results of the GARCH-MIDAS and GARCH models are better than those of the macroeconomic sentiment indices.The study finds that the financial media semantic indexes constructed by fully considering "correlated information" can improve the prediction effect of stock returns,and it is more significant in the sub-markets with relatively lower information content;(3)Stock price volatility of individual stocks takes listed companies in Shanghai and Shenzhen A-shares as the research object,and adopts negative return bias function and stock price up and down volatility ratio to portray the risk of stock collapse.It is found that financial media semantics can predict the future stock price collapse risk of listed companies,and the prediction effect is more obvious when the number of sell-side analysts tracking reports is less,the transparency of company information is lower,and the opinions of small and medium investors are more divergent.In summary,the financial media semantic index portrayed in this paper has a financial market prediction function,and the micro mechanism is that the financial media semantic reports can reduce investor disagreement and guide the market to return to rationality.Finally,this paper uses impulse response function analysis and variance decomposition to analyze the relationship between changes in financial media semantics and financial vulnerability,and finds that financial media semantics trained by two-way long-and short-term memory can have a negative impact on financial vulnerability and the vulnerability of economic and credit factors within a certain period of time,i.e.,financial media semantics can effectively reduce financial vulnerability and resolve the risk of financial vulnerability;at the same time,macro At the same time,changes in the external economic environment,the credit environment of financial and banking institutions,and monetary liquidity also have a positive impact on financial media semantics for a longer period of time;the effect of financial media semantics coverage still exists significantly after controlling for the exogenous variables that introduce financial media semantics as an influence factor on the probability of conversion.The research in this paper has very important theoretical significance and practical guidance value.In terms of theoretical significance,investors’ information acquisition behavior and information processing behavior affect financial market stability and resource allocation function,and how to obtain and identify key useful information is an important key in the information bombing era.In this paper,we use big data mining technology to obtain a semantic index of financial media based on a bidirectional long-term short-term memory neural network(Bi LSTM)model trained on news texts,and confirm that the index significantly improves the predictive power of the index by alleviating the divergence of opinions of market investors from three aspects: overall financial market,sub-markets and individual stocks,so the findings of this paper are an important addition to the existing behavioral finance literature The findings of this paper are therefore an important addition to the existing behavioral finance literature.In terms of practical guidance value,this paper portrays a semantic index of financial media that can predict financial market stability,confirms the important role of financial media reporting semantics in enhancing financial market stability and resolving financial system vulnerability,and points out that regulating the content of financial media reporting and ensuring the objectivity,independence and accuracy of financial media in news reporting can be an important tool to prevent and resolve major financial risks in China An important tool for preventing and resolving the occurrence of major financial risks in China.