Research On Information Dissemination Risk Factors In Artificial Stock Market Based On A Two-layer Network

With the arrival of information age, the status and role of informationdissemination have become increasingly important. Information dissemination ishaving a profound impact on all areas of our society. Especially, informationdissemination can lead to financial market volatility, even the financial crisis. Asinformation disseminators, investors are heterogeneous, whose microscopic behaviorare unpredictable and complicated. Combined with such factors as the complexity anddiversity of channels for information transmission, the use of traditional mathematicalmethods is difficult to describe the microscopic behavior of the market investors andinformation dissemination process. So far, many scholars have achieved someimportant results in the information dissemination research based on computationalexperiments. However, due to the limitations of the uncertainty of investors’ behaviorand the complexity of information dissemination, the existed research on the stockmarket information dissemination mechanism is not enough.For the problem of the risk formation mechanism of information dissemination instock market, this paper uses the method of computational experiments. Based on theanalysis of the results of previous studies, we summarize the characteristics ofinformation dissemination in true stock market. Then this paper build a two-layerinformation dissemination network model to study information disseminationmechanism in stock market. In the model, this paper design two types ofheterogeneous assets: stock and risk-free asset, two types of traders: fundamentaltraders and trend traders and two information dissemination networks: relationshipnetwork and media network. Also this paper consider the effect of externalinformation on investors’ investment decisions and establish the double auctionmechanism similar to the Chinese market. Meanwhile, this paper construct learningmechanism model based on traders’ total wealth.Through comparative analysis of the results of simulation and real market, wefind stock return series presents the characteristics of aiguille, thick tail and volatilityclustering. Therefore, the result verifies the model is effective. And then this paperdiscusses the way and size of risk factors impact on market liquidity and volatility.The simulation results show that the samller the market information release frequencyand the higher the information transparency, the stronger the market liquidity, and the weaker the volatility. At this time, market risk is lowest. Hover, this paper don’t getthe result of the impacts of information transmission network, the probability offeedback traders and the learning speed on stock market liquidity and volatility.Finally, according to the obtained results, and in view of the problem of informationdisclosure system in China, this paper put forward some policy suggestions.