The Research About The Degree Of Differentiation And Its Countermeasure In Credit Evaluation

Credit rating in essence should be to help investors to eliminate or reduce the mechanism of the asymmetric information, so as to provide valuable reference information of investment value and risk early warning. But in recent years the financial crisis and European debt crisis, a credit risk assessment vastly to the attention of the people, people more and more questions of credit information provided by the credit rating agency prospective and reference value. Credit evaluation should not only reflect the evaluation object, should also be convenient for market participants to classify all the evaluation object, and provide effective reference information, help them make the right investment decisions, which requires the credit evaluation has a certain distinction. So this article main content is as follows:First of all, when the distinction between multiple evaluation subject, especially in the evaluation results of multiple objects to sort or choices, evaluation of index in the process of distinguish may be considered to be "redundant indicators", or they don’t have a certain distinction. To good to distinguish between the evaluation objects, we will to filter the redundant indicators "to help distinguish between the evaluation object. How to implement effective screening process, will be one of the major problems in this paper. By using the theory of entropy index screening method is proposed in this paper, and introduce the concept of index to distinguish between degrees, from the Angle of quantitative analysis index system is optimized, which has played a role in the evaluation object. In this paper, through the high-tech enterprise credit evaluation as an example, the index system of high-tech enterprises, the optimization and selection, to distinguish between the selected optimization of high and new technology enterprise credit level played a role.In another word, because the rating agencies usually use linear methods to deal with the original data, which often leads the index with poor degree of differentiation. In a way, if the final credit scoring of the evaluated objects is with obvious degree of differentiation, standardized index score must have the degree of differentiation. So this paper will apply the S-shaped curve model. And through empirical and comparative analysis, S-shaped curve method is better than traditional linear processing method that can reflect the degree of differentiation of each evaluation subject. At the same time, it can expand the difference degree of the final evaluation results, all levels of the evaluation object get reasonable credit score. In addition, S-shaped curve processing indicator method is also an important complement to existing methods of treatment index score.