| In recent years,while the personal mortgage business of China’s commercial banks has grown continuously and rapidly,the problem of borrowers’default has gradually emerged.Controlling borrowers’ default risk to personal mortgage scientifically and effectively has become an important part of banks’ credit risk management.This thesis uses the personal mortgage data of Branch A,Bank X from 2014 to 2019,and selects 11 risk factors that affect the default of personal mortgage(including age,gender,household registration,employment stability,marital status,degree,loan amount,house area,down payment ratio,repayment period,and monthly income)from two dimensions which are personal characteristics and household financial status,and then analyze the abovementioned data empirically using Logistic Regression Model and Random Forest Model.Conclusions are as follows:(1)Borrowers with high default risk have three characteristics.Firstly,they are relatively young.Young borrowers have short employment time,more job-hopping,poor income stability,heavy living burden,few savings.Many borrowers rely on their parents for repayment.Once they encounter frictional unemployment or lose the support from their parents,it will be difficult for them to repay the loan on time.Secondly,they have non-local household registration.Borrowers with non-local household registration have poor employment stability and strong mobility,and do not have full access to social security as local residents do.They have financial difficulties covering medical expenses and their children’s educational expenses.Compared with borrowers with local household registration,their risk of default is higher.Thirdly,they are likely divorced or single.Divorced or single borrowers repay the monthly installment by themselves,and their ability to mitigate risks is low.If there is a property dispute when they divorce,failing to reach an agreement on financial arrangements as soon as possible will lead to loan default.(2)Borrowers with high default risk has four characteristics in terms of household financial status.Firstly,the loan amount is high.If a borrower gets a high amount of loan,the monthly installment is a heavy burden.If a family’s future income declines,the risk of default will increase.Secondly,the down payment ratio is low.Borrowers with a low down payment ratio have poor financial circumstances,monthly repayments and total interests are high,and they have high financial burdens.Thirdly,the repayment period is long.In a long period,borrowers may face more uncertain events such as illness,unemployment,death,and decline in income which will increase the risk of default.Fourthly,borrowers have low monthly income.Borrowers with low monthly income usually have flexible employment.They have low job stability and job security.If their incomes decline or they are laid off,their ability to repay the mortgage will be undermined accordingly.According to results of the empirical research,this thesis proposes measures to manage the default risk of personal mortgage.Firstly,enhance the development of information technology and realize the digital transformation of risk management.Secondly,improve the quality of data to ensure the accuracy of the prediction of default risk models.Thirdly,iterate default risk models and improve the whole process risk management.Fourthly,carry out mortgage policy adjustment and product innovation. |