Font Size: a A A

Empirical Research On The Impact Of Internet Public Information On SME Credit Risk Assessment

Posted on:2019-11-30Degree:MasterType:Thesis
Country:ChinaCandidate:F WuFull Text:PDF
GTID:2428330596462886Subject:Finance
Abstract/Summary:PDF Full Text Request
Small and medium-sized(SME)enterprises occupy an important position in China's national economy.For a long time,the information asymmetry between SME enterprises and financial institutions makes it difficult to assess and warn the risk of default,which forms a vicious circle of difficult financing for enterprises and bad debts for financial institutions.Although many studies have shown that financial indicators can effectively assess the credit risk of enterprises,the financial indicators of SME have low credibility,financial fraud,complex relationship and many other factors,which make it difficult to assess the credit default risk of SMEs.It is also difficult to use non-financial indicators effectively because of the lack of unified standards,difficult to obtain and quantify.In recent years,with the rapid development of the Internet in China,a large number of related information of enterprises has become Internet and open,and become an important source of non-financial information.The development of modern information technology such as cloud computing,semantic analysis and large data makes it possible to collect,collate,analyze and use Internet information data.However,there are few studies on the real impact of Internet information on corporate credit risk.This paper focuses on the Internet news information,as an important non-financial indicators into the credit risk assessment model for comparative analysis,to study whether the Internet news information can help SMEs credit default risk assessment.This paper chooses 80 listed companies in the SME edition to make an empirical analysis,and specifically studies the following contents: 1.Screening and analyzing the financial variables of listed companies.The financial data of listed companies have an important impact on the risk of corporate credit default.This paper collects the annual financial statements of listed companies for three consecutive years and extracts 26 Financial indicators.Factor analysis classifies them into six main impact factors.2.extract and quantify Internet news information.Because the news information on the Internet is scattered and all of them are text information,it needs to be crawled by crawling and mining technology and refined by text analysis technology.This paper classifies and quantifies the news information on the Internet according to the influence of media,emotional indicators,frequency of occurrence and so on,and forms the news letter on the Internet.Interest rate.3.use Logistic regression model to build credit risk assessment models and compare them.This paper analyzes the difference of the validity of the model between the pure financial indicators and the mixed indicators after adding the Internet news information,and verifies the influence of the Internet news information variables on the validity of enterprise credit risk discrimination.The results show that adding Internet news information variables can effectively improve the discriminant validity of listed companies' credit risk assessment.It shows that Internet news information can help to strengthen the identification of listed companies' credit risk,but the accuracy of discriminant decreases with the distance from default events.It is necessary to pay attention to the new Internet.Smell the timeliness of information.Therefore,if more Internet information can be considered on the basis of traditional financial indicators,it can reflect the credit status of SMEs in more dimensions,help to solve the problem of information asymmetry,help financial institutions to better assess corporate credit risk,improve the recognition rate of default risk of SMEs,and To provide references for solving the financing problems of small and medium-sized enterprises.
Keywords/Search Tags:Internet news, Text analysis, Logistic model, Credit risk
PDF Full Text Request
Related items