Based On The Discrete Choice Model Of Insider Trading Screening Study

Insider trading is an unethical behavior based on asymmetric information on the stock market, brings harm to investors and market. It’s also detrimental to the "openness, fairness and justice" principle of the stock market, thus, it’s prohibited by the regulators of every country. Although states have always been dedicated to prevent the occurrence of insider trading effectively, insider trading cases are still repeated, coupled with the complexity of the behavior of insider trading, making the regulation and law enforcement of insider trading can’t be implemented effectively. Therefore, how can we control insider trading effectively has been the focus of the studies of scholars from every country.As to this problem, we try to explore the screening model system for studying insider trading, we monitor and punish the insider trading stock timely through forecasting the probability of insider trading. First, we adopt the T-test and nonparametric tests whose mean is equal to screen out all kinds of indicator variables, and establishing discrete choice models of the first phase based on a single indicator. Finally, building the second phase of multi-factor Logistic and Probit screening model according to two approaches to in order to improve the screening efficiency of the model and provide a basis for the insider trading regulation of China’s stock market.The empirical result shows that:in terms of the screening model based on discrete choice models, the indicators and the symbols are same as expected; Screening model are closely linked with the regulatory indicators, corporate governance index and the market indicators, but the relationship with the financial indicators was not significant multi-factor screening model regard the stock equity concentration, the average daily abnormal returns and the number of the limit down as independent variables, the screening efficiency of model is90%; Screening efficiency of the Logistic model is generally the same with the Probity model among each model; the screening efficiency of the second phase model is significantly higher than the screening model based on single indicator, and the screening efficiency of the model has significantly improved; The regulatory agencies can achieve optimal regulation by controlling the significant level, to the benefit of discretion for the regulatory body; they can make the color warning according to the probability of insider trading occurrence of each stock.