Research On The Evaluation And Early Warning Of China's Financial Security

With the development of financial globalization,financial markets of various countries are closely linked and risks of international financial markets are more complex.In the current international financial situation with high leverage,high asset prices,high market volatility and high risk,the possibility of financial crisis is higher than before.From the perspective of domestic situation,economic development has entered a new normal.Economic downturn,overcapacity,narrowing interest rate differentials and the rise of Internet finance have changed the business environment of commercial Banks.The golden age of "high growth,low bad performance and high efficiency" that has lasted for more than a decade is over.The present era has entered a new stage of "low growth,high bad performance and declining benefit".The new normal of the economy is pushing the development of commercial Banks and even capital markets into a new stage.New changes will bring many disadvantages.At the present stage,financial risks will become increasingly prominent.It is of great practical significance to study financial security.This paper takes financial systemic risk as the research object,and discusses the mechanism of financial risk with the research method of combining normative analysis and empirical analysis.On the basis of evaluating the status quo of China's financial security,this paper gives an early warning of systemic financial risk and puts forward some policy suggestions to prevent it.Based on the literature review at home and abroad,this paper makes a theoretical analysis on the mechanism of financial systemic risk formation.Firstly,this paper analyzes the mechanism of financial system risk with classical theory,including the instability theory of financial system proposed by American economist minsky.Information asymmetry theory;Credit vulnerability theory;Financial asset price volatility theory.Then,the paper analyzes the mechanism of systemic financial risk in China from macro and micro levels.On the basis of theoretical research,the thesis evaluate economic development under the background of the new normal financial security situation in our country and discusses how to establish suitable for China's national conditions of China's financial risk early warning system in order to achieve early warning purposes.On the basis of summarizing the previous research results,this paper analyzes the domestic and foreign influencing factors of current financial security.Then the deep learning model the LSTM(RNN)and Google Tensorflow library tools to predict the next month values of the selects 16 indexes and synthetic index of financial safety index.Meanwhile,facebook's fbProphet industrial-level tool was used to predict the index value of each index for 6 months.Then the 16 indicators of financial security are selected to build the financial safety evaluation and early warning index system,in the past research on shadow banking is not much,add the factors in this index system(such as the off-balance sheet business in a social financing scale,makes our country finance safety evaluation and early warning system to be more perfect.In this paper,168 months of monthly data from January 2004 to December 2017 were used for empirical analysis and principal component analysis was used to synthesize financial security index.Generally speaking,the change trajectory of financial security index is in line with the historical change of China's financial security situation.On the basis of the synthetic financial security index,the risk warning model of BP neural network is used to test the result,and then the financial security situation of China in the future is warned.Then deep learning model of LSTM(RNN)is used to predict the next month and next quarter's of 16 indexes and synthetic index of financial security,using tools for Google Tensorflow library.Meanwhile,facebook's fbProphet industrial-level tool was used to predict the index value of each index for 6 months.Because deep learning technology has strong data nonlinear approximation ability and self-learning ability,it can simulate and predict complex nonlinear phenomena well,which makes up for the deficiency of neural network model.Compared with the BP neural network model of prediction result,deep learning model can keep the original data as much as possible,can widely multidimensional capture index changes,can more accurately fitting samples,warning results more reliable.To a certain extent,it provides a reference for regulators to select more effective early warning models in risk monitoring.Finally,aiming at how to effectively prevent systemic financial risks,four Suggestions are put forward:to reduce the macroeconomic levers,continue to deepen reform of the current financial regulation model,to meet the needs of mixed management reality,further improve the system of deposit insurance,improve financial ecological environment,improve the bank lending enthusiasm;On how to build a financial risk early warning system suitable to China's reality,two Suggestions are put forward:China should establish the effective comprehensive included financial risk early warning and supervision system and establish a sound financial risk early warning mechanism in our country.The innovation of this paper is mainly reflected in the following four aspects:For the first time,the application of deep learning technology in the early warning of systemic financial risk has been tested first.The research results provide a reliable basis for the establishment of financial risk early warning system in the regulatory part.The new financial security evaluation index system is based on an empirical study.This study enriches the methods of financial security evaluation and early warning.The policy Suggestions proposed on the basis of empirical research provide a practical reference for preventing systemic financial risks and constructing a financial risk early warning system suitable for China's actual situation.