Empirical Research Of Consumption Credit Scoring Model In China

With rapid development of the consumption credit in commercial bank, it is important to measure credit risk in high accuracy and predict behavior of credit applicants in future , all of that is valuable for enhancing consumer credit risk theory and improving practical management level in bank.This paper expounds the implication and classification of consumption credit firstly, and introduces the current methods of credit scoring technique. From mathematical models and comprehensive evaluation of the two angles, a combination of innovative methods is applied. This paper makes empirical research basing on actual data from a commercial bank, according to data from the sample, the variables system of consumption credit risk assessment is identified.In this paper, variables impact on individual consumer credit risk performance at this stage of China is mined through empirical research, and the performance accuracy of forecast probability from the scoring model is qualified. Firstly, the paper establishes traditional logistic regression method as an objective assessment model, next through the Analytic hierarchy process methods, gives weight of each index variable by traditional experience and actual data features, and then establishes the comprehensive consumption credit scoring model. In this way, it will take full advantages of the different methods, and also considerate the possible useful historical information, ultimate to improve the model prediction accuracy and applicability. Finally, the paper raised some of possible problems encountered at the practical application of models, and also give some suggestions about establish and improving individual consumption credit information database, which offers valuable research results about consumption credit risk assessment for financial institutions in China.