Font Size: a A A

Personal Credit Risk Assessment Method Research In China

Posted on:2017-03-07Degree:MasterType:Thesis
Country:ChinaCandidate:S S MiFull Text:PDF
GTID:2349330512959843Subject:Statistics
Abstract/Summary:PDF Full Text Request
After the founding of new China, our country has established the characteristic of the planned economy system, so as to make the credit foundation is very fragile, severely hampered the development of personal credit system. As China's reform, economic system from planned economy to socialist market economy, the consumption credit obtains the very good development, thus the credit system and its risk management is becoming more and more attention. Recently, as China's first national social credit system construction of special planning, opened a new chapter of the social credit system construction of China; At the same time in the "Credit China" web site opened in 2015, the national platform pilot project is already running on provincial access and 37 departments, the development of the social credit system makes some progress. At the beginning of the country attaches great importance to the development of the social credit system. More scholars need constantly blaze new trails and build the social credit system to lay the foundation for better.In social and economic environment promote the use of personal credit consumption is to expand domestic demand and promote economic development of the important methods. The current China's first task is to develop the economy. The use of personal credit consumption is to play the role of driving force of the national economy growth, but China's economy is still in the primary stage of socialism and the development of personal credit has encountered many difficulties, so personal credit risk is very difficult to get effective control. Th(?)fore this article research of personal credit risk assessment method is more p(?)ical significance.After reviewing the related literature, it can be seen that the study of the personal credit risk assessment for direction gradually from qualitative to quantitative direction. The quantitative analysis of the credit risk of the credit rating is the most common method of domestic and foreign research. Domestic literature often only limited to the use of Germany or Australia public credit database to have study abroad empirical improvement, credit risk assessment method for the simply consider credit evaluation method. The basic evaluation index system suitable for China's actual conditions is lack. This article adopts the research data of survey of family financial center of China as a sample data of individual credit risk assessment, a further comparative study of personal credit risk assessment model, to find more effective individual credit risk assessment model, to promote China's credit system is more healthy and rapid development.First, this article introduced from three aspects of moral credit, the credit of the law, the credit in the economy.In this paper, the current Chinese individual credit facing the main risk factors were analyzed, for example the social economic environment and lending institutions. Social and economic environment focuses on systemic risk, interest rate risk and risk of policy and law. We refers to the lending institutions mainly commercial Banks. From the lenders, at present, the main risk including personal credit risk, liquidity risk, operational risk, etc. Lenders are one of the most important risk of individual's credit risk. Personal credit risk is mainly manifested in the debtor defaults, the borrower's credit rating changes, etc. The current personal credit risk are mainly as follows:the borrower's performance ability lower, the borrower's repayment willingness to fuzzy, false mortgage. The current operational risks are mainly concentrated in:bank loan eligibility criteria has decreased, bank management system is not perfect, technical level is relatively backward, lack of legal basis. Liquidity risk mainly refers to the current commercial bank assets and liabilities "date wrong "-"save short and lending long" phenomenon, resulting in a funding liquidity risk.Second, the individual credit risk assessment process is divided into the following four parts:(1) Definite the problem(2) Collect and pretreat the sample data;(3) Establish personal credit risk assessment model;(4) Do the model test, interpretation and application;This article when the quantitative method of credit risk management in detail, such as expert criterion, logistic regression, decision tree, neural network model, and compares their advantages and disadvantages.Three, according to China's national conditions and the personal credit risk of Commercial Banks at home and abroad for reference, the assessment index system, finally selected 24 individual credit risk assessment indicators. We will through the method of quantitative analysis for more than 24 indicators of primary measure of the personal credit risk identification ability. According to the quantitative criteria for further index screening, eventually we established a simple and effective system of personal credit risk assessment.In this paper, the individual credit risk assessment index of the ability to recognize the discriminant:(1) the independent sample t-test, six evaluation indexes recognition ability of the personal credit risk is poorer, So we need to delete marital status, other non-financial assets, checking account deposits, regular account deposits, holding cash, obey the traffic rules the six evaluation indexes from the personal credit risk assessment index system.(2) Wald test or nonparametric statistical test,5 indicators to identify individual credit risk ability is poorer. So we need to delete marital status, other non-financial assets, checking account deposits, regular account deposits, obey the traffic rules the five indicators from the personal credit risk assessment index system.Four, in paper logistic regression statistical method is further subdivided into Forward Stepwise logistic regression and Backward Stepwise logistic regression. Logistic regression model is applied to the individual credit risk assessment.Based on Forward Stepwise logistic considering from the perspective of personal credit risk management, need special attention in the personal credit risk assessment index system:year after-tax money wages, credit card record, have to apply for loans in the bank number, housing situation, professional and technical titles, political affiliation, stock account.According to the Backward Stepwise logistic regression, considering from the perspective of personal credit risk management, in the individual credit risk assessment index system should give special attention:year after-tax money wages, credit card records, job preparation, has filed in the bank loan project, the agricultural registered permanent residence, housing situation, the political landscape, stock account.Five, in order to better evaluate the credit risks of the personal credit judgment personal, we try to logistic regression analysis method and clustering analysis method. This paper adopted based on the clustering of logistic stepwise regression analysis of hybrid method to construct personal credit risk assessment model.The logistic regression model regression to identify clustering factors, moreover the recent distance method is adopted to classify sample data, finally to realize the effective classification of personal credit; At the completion of the establishment of the comprehensive personal credit risk assessment model, by using ROC curve on the model for further inspection.We use SPSS software using the maximum likelihood method logistic stepwise regression method, eventually filtered identified nine indicators. Using clustering analysis further confirmed the four factors as the political landscape, culture degree, housing situation, credit history. Eventually establish bilateral clustering model of logistic regression models were compared with bilateral cluster statistical model concluded bilateral cluster statistical model is more effective.Six, Conclusion and shortage.
Keywords/Search Tags:personal credit, credit risk, Logistic regression model, clustering analysis
PDF Full Text Request
Related items