| With the occurrence of many financial crises in recent years and the acceleration of economic globalization,the global financial environment has fallen into a complex and changeable situation,and the economies of all countries will encounter many difficulties and challenges.In addition,the sudden outbreak of COVID-19 in 2019 has further exacerbated the imbalance of the world economic order,and the world economy is in an unprecedented difficult period.As a major branch of China’s financial system,commercial banks face such an economic environment,the identification and early warning of financial risks has become the focus of daily operation and management.At present,constructing a scientific and effective risk early warning model and constantly improving the financial risk early warning mechanism are the urgent and important core content of internal supervision construction for China’s commercial banks.Domestic scholars should learn from foreign research experience in the construction of financial risk early warning mechanism,and should fully combine the current domestic economic development level and China’s banking supervision mode,build a financial risk early warning system in line with China’s commercial banks and formulate the financial risk control mechanism.This paper selects the quarterly data of 40 A-share listed commercial banks in China in recent years as the sample data,and constructs the financial risk early warning index system of Chinese commercial banks according to relevant regulatory regulations and relevant research literature.The selected sample data has a strong representative industry,the financial risk early warning index system from multiple perspectives,selection of assets,net interest rate,cost-income ratio,net profit margin and financial indicators such as non-performing loan ratio,single largest customer loan ratio of banking financial indicators,system comprehensively reflect the main financial risks facing the financial Banks with less financial indicators facing the financial risk.The correlation analysis of the early warning index system is analyzed,and the scatter chart and heat map are used to output the analysis results intuitively.Based on the analysis results,the principal component analysis method is used to extract the principal components and determine the index weight,and quantify the comprehensive risk value of each individual bank in each quarter,so as the basis for dividing the financial risk level.The SVM model in machine learning is selected to give financial risk early warning to each individual bank.The research shows that the financial risk early warning model constructed based on SVM model has high accuracy and can be used in the financial risk early warning work of Z Bank.The financial index data of Bank Z in recent years were brought into the risk early warning,and the accuracy test and effectiveness of the early warning results were analyzed to further verify the applicability and reliability of the constructed model.Finally,the early warning results are analyzed,and the corresponding risk prevention measures are put forward for the targeted problems found,so as to improve the reference significance for the use of Bank Z in avoiding financial risks in the subsequent operation process. |
