Stock price manipulation forecasting is to carry out judgment before the stock price being manipulated, and to identify indicators with which the stocks are likely to be manipulated. Based on the 2015 A-share market stock trading manipulation case panel data issued by China Securities Regulatory Commission, and on the basis of two facts that financial advantage is needed for a manipulator to implement manipulation, and that other investors are permitted to copy trading, the panel data Logit model is built to forecast the stock price manipulation by identifying the main influencing factors of stock price manipulation with the stepwise regression method. The results show that when the average stock holding ratio and the price volatility of the stock in the previous quarter are higher, and the beta value is lower, then it is more likely for the stock price to be manipulated. The overall prediction accuracy of the out-of-sample and the in-sample is 68.18% and 70.63% respectively, so the effectiveness is relatively good.
