A Statistical Analysis Of Bank Credit Risk Based On Economic Model

The problem of non-performing assets has plagued banks for a long time and becomes a major financial risk for the banking industry,directly threatening the survival and development of banks.Among the risks of the bank,credit risk is the most prone to happen.Therefore,credit risk management has become a top priority in bank management.Some banks will adopt extreme and negative measures to prevent credit risks.For example,customers who subjectively under suspicion that they are in breach of contract will be refused of their loan requirements by the bank.This will not only fail to fundamentally solve the problem,but will even cause the bank to lose a large part of its quality customers.In order to strengthen the bank's credit risk management and meet customer's requirements in the loan business,effectively reducing credit risk and non-performing loans,the thesis applies some methods in data mining to the bank's credit risk research and application,combined with economic theory.The experiments proved the effectiveness and scientificity of the overall credit assessment;and based on the experiments results,we put forward some operative suggestions.Specific content includes:(1)Introduce the statistical methods that will be used in the thesis,including the random forest algorithm and principal component analysis method,as well as theoretical models in economics,such as the RFM(Recency?Frequency?Monetary)model and the STP(Segmentation?Targeting?Positioning)theory.(2)Based on the advantages of random forest and the characteristics of the RFM model,the two are combined.Through the experimental comparison,it is found that the accuracy is improved and the time taken is also reduced.(3)Use the advantages of principal component analysis to identify the most important variables in credit risk,then analyze and apply these variables using STP theory,especially the key factors,and make decisions to reduce credit risk,so that credit risk can get Maximum control.