| With the development of the Chinese economy,Mergers and acquisitions have become one of the effective ways for enterprises to expand their scale,improve their competitiveness and realize diversification.However,since the domestic market mechanism is not sound enough,and the development time for enterprise M&A is short,there still exist many problems in the area.Blind expansion,excessive government intervention,tax avoidance,insider trading and other phenomena emerge endlessly.To protect the interests of ordinary investors,it is of great importance to reduce the risk of investment through quantifying the stock price after M&A.Good results have been obtained in solving the quantitative forecasting of stock price by using the traditional Time Series Model,ANN model and etc.,but these models have their own limitations.SVM model has unique advantages in solving problems like over-learning,less learning,local minimum,curse of dimension,so on and so forth;therefore,it has been widely employed in short-term stock price forecasts.This paper focuses on the M&A cases of Alibaba,Tencent and Evergrande Real Estate Group and uses the SVM model to make short-term predictions of the stock prices after M&A.Firstly,this paper introduces the background of M&A and the relevant existing research literature,summarizes the main factors influencing stock price fluctuation in M&A,and discusses the basic theory of support vector regression machine in details.Secondly,we use the support vector regression machine in the present paper for predicting the stock price after M&A and constructing the basic procedure of price forecasting,and build a database based on the M&A cases of Alibaba,Tencent and Evergrande to fully delve information about M&A.In order to further improve the prediction accuracy of the model,SVM share-price quantitative analysis prediction model based on variable selection is established.We first select all the factors that may influence the stock price fluctuations and then use gray relational analysis to screen the influencing factors.After selecting the selected indicators as input variables,and the share price as the output variable after the purchase of the case,we select Gauss kernel function,use grid search method to find the optimal parameters.The empirical results show that the support vector regression based on variable screening can achieve a good fitting effect on the short-term share price after M&A. |
