Research On The Factors Influencing Non-Performing Personal Housing Loans In P Branch Of G Bank During The COVID-19 Pandemic

Personal housing mortgage loans are typical personal credit products offered by commercial banks.With the development of the times and the gradual maturity of China’s real estate market,more and more residents are choosing to use mortgage loans to purchase houses,repay monthly payments with future income,and ultimately obtain ownership of the house after repaying the loan.Commercial banks are the main lenders of mortgage loans.As the scale of mortgage loans gradually expands,commercial banks are constantly exploring how to manage the risks of this part of assets.Since the outbreak of the COVID-19 pandemic,the asset quality of commercial banks’ mortgage loans has gradually come under pressure due to the increasing number of overdue loans that borrowers cannot repay in a timely and full manner.This paper explores which factors have a more significant impact on borrowers’ default and how to identify and prevent these factors.Based on the above content,this paper first sorts out the research literature on housing mortgage loans,non-performing loan risks,and COVID-19-related issues at home and abroad.On the basis of the theory of information asymmetry,the relevant concepts of housing mortgage loans and the causes and risk control of non-performing loans are sorted out.The current situation and macro background of commercial banks’ non-performing loans in China are analyzed,followed by a review of the impact of the COVID-19 pandemic on bank loans.Then,taking P branch as the research object,by sorting out the basic information and loan information of borrowers,factors such as borrowers’ gender,age,marital status,education level,loan amount,and down payment amount are selected as observation variables.After comparing multiple econometric analysis methods,logistic regression and random forest methods are used to analyze credit default risk variables and construct a suitable model to explore the factors that affect the default of P branch borrowers.In the empirical analysis,the Logit regression method was used to select variables including gender,education level,age,loan term,and personal annual income as significant features.Among them,gender(male)and loan term were positively correlated,while education level,age,and personal annual income were negatively correlated.According to the importance feature ranking results of the random forest,the contract loan balance amount ratio,loan term,down payment amount,personal annual income,age,and education level were ranked at the forefront,which was consistent with the results of the Logit model and the model output was stable.This article proposes targeted risk control suggestions for significant influencing factors,such as conducting comprehensive information collection of borrowers before the loan,using risk screening tools and data models of commercial banks to strengthen review during the loan process,continuously monitoring new factors identified in the research process,such as the "contract loan balance amount ratio" index.Meanwhile,it is recommended that commercial banks use external insurance to reduce credit default risk.