Construction Of The China's Mixed Financial Safety Warning System

After China's entry into WTO, the opening-up of our financial sector will be deepened in the future. In the meantime, the development of the financial internationalization will cause increasing risks that our country may be confronted with. It has been the study focus of modern financial safety theory on how to do well with early warning of financial safety to guard against financial crises to ensure the safety of the Chinese financial system.This thesis first introduces and analyzes several methods that are widely employed by experts both at home and abroad, and then it indicates the advantages and disadvantages of these methods. On the basis of these results, it brings forward the basic theorem and the process of the mixed financial safety early warning system.Secondly, in terms of financial safety situation that China is facing under the condition of opening-up, it illustrates the internal relation among financial safety, financial crises and financial risks. Then, it puts forward and analyzes internal and external factors that would exert influences on financial safety.Thirdly, on the basis of the analyses about China's financial risking factors as well as combined with the research results both at home and abroad, this thesis has collected fourteen indicators, which are divided into four subsystems including macro financial safety subsystem, banking safety subsystem, currency safety subsystem and stock-estate safety subsystem, to build up financial safety early warning indicator system. After that, it employs the Entropy method, inputting data from the year of 1994 to 2002, to reckon these indicators' weight on the overall system to overcome faults of traditional methods. As a result, it paves way for the further research.Lastly, in order to overcome weaknesses of traditional early warning methods, back-propagation neutral network (BPNN) is introduced into financial safety warning system and is connected with indicator warning method to build up BPNN warning subsystem facing to MATLAB and indicator warning subsystem. On the one hand, it can utilize the merits of nonlinear modeling, fast self-studying and strong extending of BPNN; on the other hand, it takes use of the merit of indicator warning method that has strong explaining ability to make up the deficiency of BPNN. With the employment of the combination of these two methods, the mixed financial safety early warning system can predict the safety situation of nonlinear financial system as well as explain the output of BPNN subsystem in accordance with the true situation in our country.The mixed financial safety early warning system that this thesis attempts to establish has been realized on the computer through some soft wares such as MATLAB, SPSS, etc. This thesis first takes statistics...