Research On The Financial Distress Prediction And Control Measures Of Chinese Non-financial Enterprises

As the main value creator of the national economy,non-financial enterprises play a vital role in economic development.Since the financial crisis in 2008,the debt ratio of non-financial enterprises in China has continued to grow,reaching its peak in 16 years.In recent two years,although the government has paid attention to preventing and resolving risks,implemented supply side reform and "deleveraging" policies,and used macro-control to reduce the debt ratio of non-financial enterprises,the debt ratio of non-financial enterprises in China is still high.Under the downward pressure of economic cycle,the wave of non-financial enterprise debt default broke out in Zhejiang Province and Shandong Province.It can be seen that the financial risk of non-financial enterprises needs to be focused.It is usually a continuous process for enterprises to fall into financial difficulties.In the process,companies themselves will inevitably show signs of deterioration in their financial conditions.By observing the changing trends of these symptoms,corporate stakeholders can enhance the company's monitoring capabilities and take measures to curb the deterioration of the financial situation.Therefore,early warning methods that comprehensively reflect the financial status of the enterprise in a timely manner are particularly important for stakeholders.Reviewing the literature at home and abroad,the study of financial distress has attracted much attention both in practice and in theory.In the past few decades,experts at home and abroad have mainly used financial information indicators,cash flow information indicators and market yield information indicators to build early warning models.However,most of the existing researches on the debt crisis of non-financial enterprises are at the macro level,lacking in-depth micro analysis.In this paper,combined with the current macroeconomic situation,taking into account the factors outside and inside the balance sheet,with the help of logistic regression analysis with MCP penalty term and Bayesian discriminant analysis,we build a financial distress early warning model for non-financial enterprises.The purpose is to find out an efficient and accurate early warning method,study the relationship between changes in financial indicators and financial distress,and then provide timely and effective policy suggestions.The innovations of this article are as follows:Firstly,the article uses unbalanced data to establish an early warning model,and uses ROSE algorithm to synthesize artificial data,which makes the data more suitable for the actual economic situation and reduces the prediction error caused by oversampling of positive samples;Secondly,under the pressure of the current economic downturn,this paper comprehensively considers the on-and off-balance sheet impact variables,and includes the external guarantee as a contingent liability in the model;Thirdly,based on T-2 and T-1 financial data,this paper builds an early warning model with the help of Logistic regression analysis method and Bayesian discriminant analysis method with MCP penalty,and explores the influence of the sample data interval on the prediction accuracy of the model And the advantages and disadvantages of the two analysis methods.The conclusions of this paper are as follows:Firstly,compared with T-1,using financial data of T-2 to establish early warning model can effectively improve the prediction accuracy;Secondly,through comprehensive comparison,the prediction accuracy of logistic regression model with MCP penalty in T-2 is the highest,which is the optimal model;Thirdly,the results of model fitting show that variables X1(ROA),X9(asset liability ratio)and X11(The ratio of external guarantee to net assets)has a significant correlation with financial distress,and has a strong ability of explanation.