Researches On Computer Simulation Design Of Rumors’ Effects On Stock Price

In the present securities market of China,most of the individual traders are irrationally treating stock trading and hold risk-taking mentalities,so that they are sensitive to various kinds of outside information.However,in order to gain higher profit,professional investment institutions or other stakeholders,often spread distorted or error information(i.e.rumors)by Internet medias to mislead the investing public,resulting in the distortion of stock trading volume and the abnormal fluctuation in stock price.Due to the diversity of commercial information,diversity of investors and complexity of stock price fluctuation mechanism,the rumor impacts on the stock price cannot be simply derived by mathematical theory through importing simple assumptions.Moreover,since in the real stock trading,there has not been any effective technology to distinguish the trading of individual investors and professional institutions,the rumor influence on stock price cannot be judged through the analysi s of the real trading data.Multi-agent-based simulation technology developed rapidly in the past 40 years,and has been widely used in many studies of commercial activities since the 1990 s and gave many satisfactory results.To study the effects of commer cial rumors on stock trading,the research based on multi-agent simulation technology for the rumor spreading and rumor-related stock trading is indispensable.Employ the latest achievements in the fields of rumors spread research,artificial intelligence research and computer science,this thesis constructs individual agent on the micro scale,including micro individual attributes,behavior rules and learning ability.By simulating the rumor spreading,the individual’s response to rumors,and transactions with individual game,the thesis deduces macro rumor effects on stock trading,which is a "bottom-up" research methods.Focus on the securities market,this thesis firstly gave the mechanisms of rumor generation,spreading,extinction and evolution,in which,the types and generating time are decided by the profit and stock amount of the professional investment institutions,the spreading of rumor is decided by the connections of individual traders,the extinction of rumor is decided by the reactions of individual traders and the rumor life cycle,the evolution of rumor are determined by the gaming between rumors and traders.This thesis then defines a variety of agents,including rumors agent,individual investor agent,institutional investor agent,the price forming agent,the environment agent,the network connection agent,etc.,and defines the properties and the behavior rules of these agents,through which we realize simulation of the reaction processes of investors to rumors,trading game,evolution of trading strategy,and the stock price fluctuation.Finally this thesis construct a rumor-related stock trading simulation framework to implement the simulation process.In simulation of this thesis,through reasonably configuring parameters,the stock price fluctuation and stock trading amount variation with commercial rumors have been obtained.We analyzed the impaction of rumors by combing the simulation results and stock trading in reality.The study of this thesis may shed some light for future study of stocking price fluctuation mechanism.