Borrower’s Behavior Information Of Mobile Phone And Consumer Finance Default Risk

In recent years,massive financial data has exploded with the transformation of artificial intelligence technology and financial technology innovation,making credit data complex,diverse,and heterogeneous.Traditional financial data analysis methods cannot effectively evaluate personal credit.It leads to a high default risk of borrowers in the market,and even malicious default,which has a negative impact on the sustainable and healthy development of the Internet consumer finance industry.Since the customer groups of Internet consumer finance are mainly small and medium-sized enterprises and low-income earners who cannot provide collateral,it is essential to solving the information asymmetry problem between borrowers and lenders.Nowadays,smartphones have become an indispensable part of people’s lives,containing rich personal information,such as communication records,text messages,network footprints,standard addresses,and APP installation information.This information is objectively and automatically recorded by the mobile phone system,with minor measurement errors,and is closely related to personal life.It can reflect the personality and demographic characteristics of mobile phone users.Therefore,the daily data of mobile phones have great potential in identifying personal credit default risks.We collect the information of more than 700,000 borrowers and some borrowers’ mobile phone behavior information from the Indonesian Internet lending platform A,and systematically and comprehensively discuss the relationship between the borrower’s mobile phone information and the borrower’s default risk under the assumptions of different behavior theories.Since the data comes from different channels and experiments of the lending platform,it is impossible to fully obtain all the information about each borrower’s mobile phone.Based on the information that can be obtained,the ability to identify the borrower’s default risk is tested from four perspectives and theories.Specifically,the research contents and conclusions of the paper are as follows:(1)Based on the borrower’s social capital perspective,we use the borrower’s address book information to test its role in identifying the borrower’s default risk.The study found that the more contacts in the borrower’s address book,the lower the risk of loan default.Further,we extract relatives’ nickname address book information through the nickname information of the borrower’s address book.It is found that the more relatives’ nickname address book is,the smaller the default risk is.In the heterogeneity analysis,compared with borrowers with a bachelor’s degree or above,the social capital of borrowers with a bachelor’s degree or below is weaker in identifying their default risk.(2)Mobile APP information is closely related to personal life,and different types of APP installation information reflect the personality and preferences of mobile phone users.Based on the borrower’s happiness preference perspective,the borrower’s APP is classified into realization-type APP and hedonic-type APP,and its role in identifying default risk is tested.The study found that: the more APPs installed by the borrower,the greater the default risk;the more the borrower has installed APPs,the lower the default risk;the more hedonic APPs are installed,the greater the default risk.Further,we draw the same conclusion from the relative quantitative analysis of the two APP categories.Regarding heterogeneity,from the perspective of gender,the total number of APPs substantially affects the default identification of males.However,realization-oriented APPs have a better effect on the identification of default risk of female borrowers.In terms of age,the realization preference APP can better identify default risks in borrowers over 35 years old.After that,the relationship between APP dynamic installation information and default risk is analyzed,and it is found that between two loans,borrowers who prefer realization-type APPs have lower default risk.(3)In the measurement dimension of perceived value,the shopping time of consumers is an important reference dimension.Based on perceived value,we study the relationship between the borrower’s loan application time and default risk.The conclusion shows that the shorter the borrower’s loan application time,the higher the perceived value of the loan,and the lower the default risk.In the discussion of heterogeneity,the loan application time of the borrowers in the high-interest rate group has a weaker identification effect on the default risk,whether it is analyzed from the significance or the size of the coefficient.After dealing with endogenous problems and robustness tests,it is still shown that perceived value affects individuals’ behavioral intentions.As a special commodity,borrowers’ perceived value of loan products can effectively identify borrowers’ default risk.(4)The relationship between the borrower’s mobile phone number history,application loan information,and default risk is analyzed based on the perspective of behavior and habits.The study found that the more borrowers have historically applied for loans in the entire lending market,the greater the risk of default.When classifying the lending market,the more borrowers have historically applied for loans in the high-quality lending market,the lower the default risk.The more historically borrowed applications in the low-quality lending market,the greater the default risk.At the same time,we also examine the concentration of loan history applications.The results show that whether in the high-quality or low-quality loan market,if the borrower has concentrated borrowing behavior in the near future,the loan also has a higher risk of default.In the discussion of heterogeneity,it may be affected by the epidemic.During the epidemic,the borrower’s borrowing concentration and high-quality loans have weakened the ability to identify default risks.In the dynamic change analysis,the results show that the borrower’s default risk will increase between two borrowings,whether the borrower has more high-quality loan applications or low-quality loan applications in the external lending market.Different from existing research,the innovation and contribution of the paper are mainly reflected in the following points:(1)The paper enriches research on the default risk of Internet lending in emerging economies based on a relatively unique piece of data.Existing research on Internet consumer finance mainly uses relatively developed countries as research samples,and there are few types of research on emerging economies in Southeast Asia.In different cultural backgrounds and economic environments,the default risk characteristics of borrowers are inconsistent.Based on previous research,we take Indonesia,the fourth most populous country in the world,as a research sample.It combines the microscopic behavior data of borrowers to test the relationship between Indonesian borrower behavior information and default risk.A valuable addition to the fintech literature in emerging economy markets.(2)This study contributes to supplementing the relevant literature on factors for identifying borrower default risk.In historical documents,the borrower’s demographic information,financial information,platform certification information,social capital information,and other characteristics can effectively identify the borrower’s default risk and display the borrower’s credit signal.Based on relevant research,we explore the relationship between borrower behavior information and default risk from the perspectives of borrowers’ social capital,happiness preference,perceived value,and historical lending behavior,combined with borrower mobile phone behavior information.The borrower’s address book information complements the research on social capital and default risk.At the same time,innovatively identifies borrower default risk from borrowers’ happiness preference,perceived value,and historical lending behavior,providing different perspectives for borrower default risk identification.(3)In the historical literature,most of the research analyzes the relationship between the borrower and the default risk through the borrower’s static information,such as the borrower’s financial income,occupation,and so on.Nevertheless,some characteristics of borrowers can change over time.In related research,few scholars have studied the relationship between the dynamic changes in borrowers’ behavior and default risk.We supplement the influence of the borrower’s behavior change on the default risk during the two loan periods through the borrower’s multiple loan information on the platform.We innovatively study the dynamic changes of the borrower’s behavior,that is,the influence of the borrower’s behavior changes on the default risk during the two loan periods.