The Impact Of Demographic Factors On The Risk Of Short-term Loan Default

Personal short-term loans refer to personal loans with a term of less than one year(including one year).In January 2015,the amount of personal short-term loans was only 8.18 trillion yuan.By October 2018,the balance had increased to13.43 trillion yuan.In recent years,the scale of personal short-term loans has grown rapidly,increasing by nearly 50% in the past three years.It has become an important part of personal loans in China besides housing mortgage loans.With the rapid growth of loan scale,it can not be ignored that the non-performing loan rate is growing rapidly.By October 2018,the non-performing loan rate was about1.87%,a 0.31 percentage point increase over the same period last year.Therefore,the default risk of personal short-term loans has become a realistic issue worthy of study and discuss.Based on the demographic characteristics of short-term loan borrowers,this paper studies the default risk of short-term loan,analyses the theoretical and empirical relationship between borrower's demographic characteristics and default,explores the influencing factors of short-term loan default from the perspective of borrower's demographic characteristics,and uses borrower's demographic characteristics to identify and predict the risk of short-term loan default.Firstly,based on the analysis of the current situation of personal short-term loan default risk,this paper studies the relationship between the borrower's demographic characteristics and personal short-term loan default risk from the perspective of the borrower's demographic characteristics.By dividing the characteristics of sex,age,marriage,education,occupation and industry of borrowers,this paper explores the relationship between natural and social attributes of borrowers and short-term loan default.The study finds that female population,married population and highly educated population have relatively low default rate of short-term loans;relatively,male population,divorced population and low educated population have higher default rate;there is no correlation between different industry characteristics and age characteristics of borrowers and their short-term loan default,and the company-type form of professional organization has loan default rate than other types of borrowers.Empirical research results have two enlightening meanings for credit risk management of commercial banks.First of all,we should pay attention to the risk review of demographic characteristics.We should not only pay attention to the borrowers' repayment ability in the sense of economic indicators,demographic characteristics of borrowers can also reflect default risk.For example,the borrowers' marital and academic characteristics will reflect the repayment intention and ability from one side.Differential review and credit scheme should provide according to the demographic characteristics of borrowers;secondly,commercial banks and other credit institutions should not adopt biased review method for the age and industry factors of borrowers,nor should they adopt conservative risk control measures when the industry is facing cyclical fluctuations,so as to reduce the amount of loans to the corresponding industries.In order to improve the ability of controlling the risk of short-term loan default,we try to study effective risk control methods.Using the empirical results,this paper builds a stochastic forest model by introducing demographic and loan factors with significant characteristics,in order to predict and judge the individual short-term loan default.By using the restored RUSboost random forest model to predict the typical unbalanced samples of short-term loans,it is found that the combination of demographic characteristics and basic loan characteristics has explanatory and predictive effects on short-term loan default.As far as the sample is concerned,by using demographic characteristics to predict the default of short-term loans,the accuracy rate of default recognition is about 82%.Through the stability detection of the results,the recognition rate of prediction in short-term data of online loans can reach about 80%.Based on the predictability of short-term loan default due to demographic characteristics and loan factors,the credit scorecard system based on demographic characteristics and loanfactors is established,and the results of this study are applied to commercial banks.Through the establishment of credit scoring system,the impact of demographic characteristics and loan characteristics on short-term loan default is visualized,and a rating table for borrowers is established.By calculating the default probability of each scoring range,the risk probability comparison table(the main yardstick)for short-term loan default is obtained.Therefore,in the practice judgment of short-term loan default,the feasible operation scheme of basic prediction based on demographic characteristics is established.Through theoretical analysis and experience summary,it is found that the internal factors of demographic characteristics affecting loan default can be divided into two parts: repayment ability and repayment willingness.The reason for short-term loan default is the result of the comprehensive struggle between borrower's repayment ability and repayment willingness.It is difficult to fully evaluate borrowers by collecting the characteristics of traditional demographic factors using traditional demographic methods.In view of this,this paper makes a trend discussion on the impact of short-term loan default based on the big data of the Internet.we believe that the collection of information based on population behavior data will accurately describe and characterize the borrower's repayment intention in the future,and that a single populationmay transform to a digital individual in the foreseeable future.Credit evaluation and loan default correlation research will enter a new era.At that time,the correlation analysis of demographic factors on short-term loan default will expand more effective variables.These variables,whether demographic or non-demographic,will make the observation of default behavior more comprehensive and accurate,and the prediction of loan default will achieve better results.