An Empirical Analysis And Discrimination Research On Insider Trading And Market Manipulation In Chinese Security Market

Insider trading and market manipulation belong to security fraud, which the authorities all over the world supervise because of their great harmfulness. Insider trading and market manipulation are universal phenomena, and because the means used are generally covert, it is very hard to supervise them. Therefore how to regulate insider trading and market manipulation effectively is an important problem in the developing of our security market. So this paper introduces probability neural network to set up a model to discriminate the insider trading and market manipulation in Chinese security market. First of all, for providing empirical evidence to the selection of the model's input variable, this paper should research on the impact of insider trading and market manipulation on Chinese security market in the sight of operational performance, such as stock liquidity and prices. There is little research on the impact of insider trading and market manipulation on Chinese security market operational performance. Furthermore Chinese security market is very different from other markets. So this paper research on operational performance, such as stock liquidity and prices, to measure the impact of insider trading and market manipulation on Chinese security market. Based on it, we set up a model with probabilistic neural network to discriminate the insider trading and market manipulation in Chinese security market.First of all, this paper calculates the character of market operational performance, such as stock liquidity and prices, based on the sample of insider trading and market manipulation supplied by CSRC. The result shows that, firstly, in Chinese security market, during insider trading and market manipulation, stocks became more liquid. We consider that it is because the proportion of individual investors is high and the means in the trades are covert. Secondly, market maniplution is effective, and insider trading has manipulation effect on market to a certain extent. We also find empirical evidence indicating that in Chinese securities markets, insider trading reveals news in advance, so it increase the stock market's speed of adjustment to news.Secondly,based on the empirical research, we try to set up a model with probabilistic neural network, and use it to discriminate the insider trading and market manipulation in Chinese security market. The result shows that the model set up in this paper performs well. Compared with Logistic model, its discrimination accuracy is much higher. Furthermore, it is easier to design.Finally, based on the conclusion and the realities in Chinese security market, this paper indicates the problems existing in the supervision of insider trading and market manipulation in Chinese securities market and gives some suggestion on the financial supervision in China.