Financial Crisis Pre-warning Model Of Listed Companies Based On The BP Neural Network

Along with the economic globalization and the continuing reform of China’s politics, economy and culture, China’s financial markets have developed rapidly and tried to catch up with the whole world. The listed Corporations that raise funds to develop and operate have become the main body of the securities market, whose operation has drawn much attention of all the stake holders. It seems that China’s securities market has a thriving scene, but the overall operating performances of listed companies are not optimistic, quite a few listed companies suffered a lot. One company can hardly survive the financial crisis under the negative affection of the poor management level and the down-going market. Once listed companied come across with financial risks, it will affect China’s capital markets and the interests of stakeholders. In this case, in order to help the authorities to implement effective monitoring policy, to help business owners prevent crisis, the establishing of a reasonable, accurate and real-time effective financial crisis pre-waraing model is extremely necessary.Current methods for early warning model of financial crisis can be summed up as the traditional statistical method and artificial intelligence method. The advantage of the former method is simple, intuitive and enjoys specific analytical formula. However, the former one is restricted in variables, namely the statistic model needs to assume the variables to meet a series of statistical characteristics. But that is only a special case, which is out of touch with reality and unconvincing. Shortcomings of traditional statistical methods make it defective in data fitting and model predictive accuracy rate, which make the model lack of wide spreading availability. Although the neural network does not have any analytical formula and the training process of it is more abstract, as it can make up the shortages of traditional statistical methods, it has been widely used in various fields. BP neural network is one of the most widely used neural networks.In this paper, the author makes a fairly investigation of the literature about financial crisis both home and abroad, and then makes a comparative study of the financial crisis warning techniques using various ways with the aid of a precise definition of the financial crisis. Hence, the author recognizes the shortcomings of traditional statistical methods and advantages of non-statistical methods. Then comes the empirical study, through comparing the results of the financial crisis early warning model based on neural network model and Logistic regression, the study finds that neural network model is more precise.The innovation lies at its relatively new angle of this paper. Although researches of crisis warning neural network already exist, this article uses the latest data, and the data are mainly drawn from the manufacturing industry; this article also makes a more stringent definition of the outcome given by the computer; and through comparing with the traditional statistical methods, the privilege of the BP neural network prediction model is elaborated.