The Study On Network Public Opinion Warning For Listed Companies By Using The Financial Characteristics

In recent years, the network public opinions of listed companies have grown rapidly, which lead to the company’s share price fell, damage the benefit of shareholders and so on. Therefore, building an effective early-warning system of the network public opinion of listed companies is very important. Early-warning of network public opinion produces early-warning signals at the beginning of the network public opinion, which prompts companies go on alert and then take appropriate steps to protect prevention, and then response to crisis caused by negative network public opinion, which is the important part of corporate governance and internal control.This paper analyzes the evolution of the network public opinion event by using the information asymmetry theory, the efficient market theory and the behavioral finance theory and clarify the definition of the network public opinion of listed company, then takes the listed company in Public Opinion List of 2012 as research subject to determine the index system. At last, the paper models a research system of negative network public opinion of listed companies based on BP neural network. Because of the establishment of the early-warning model, business managers will be able to take effective measures to avoid or reduce the damage of the crisis. By the research model, we found that in the media, institutional investors trade number, and so on are occupying very important position in the negative network early-warning system of listed companies, the importance in the process of the early-warning model is close to 9%, and accuracy of the model can up to 75%, which achieved good prediction effect.Through the establishment of early-warning model of listed companies, we try to improve the system of internal control of listed companies, to provide powerful tools for the implementation of effective internal control, and to achieve the effect of warning and against early.