The Research Of Small Enterprises Credit Rating Based On Distinguished Default Significantly

The research of small enterprises credit rating is to establish a credit rating model which can significantly distinguish the default state of small enterprises. The research of small enterprises credit rating is composed of two parts, the establishment of the small enterprise credit rating model and the empirical research.This paper is divided into four chapters. The first chapter is the introduction, this chapter expounds the features of the small enterprise credit rating, besides, the present situation of small enterprises credit rating and the existing research problems are discussed. The second chapter is to establish a credit rating model for small businesses, including the construction of the small enterprise credit rating index system, the construction of the credit rating model and the construction of credit rating system. The third chapter makes an empirical analysis on a sample of 3111 small enterprise, builds a small enterprise credit rating system which can significantly distinguish the default state. The fourth chapter summarizes the main work and the innovation of this thesis. The main works of this paper are as follows:(l)This thesis establishes a small enterprises credit rating index system which can significantly distinguish the default state. First, divide the data of index i of all small enterprises into two kinds of samples, default and non default. According to the idea that ratio of the group dispersion and total dispersion of index i is smaller, the ability of index i to discriminate the default state is stronger, select the index with the strongest ability. Then F test is performed on this index. If the index went through the significant test, bring the index into the index system, if not, delete it. After bringing a new index into the index system, test the ability of the indexes had been brought into the index system one by one. If the discriminate ability becomes non-significant due to the introduction of new index, it should be deleted from the index system. Repeat the above steps until the indexes which can significantly distinguish the default state have been brought into the index system. Next, conduct partial correlation analysis on the rest indexes and exclude index containing redundant information. Thus, a credit rating model which can significantly distinguish the default state of small enterprises is established.(2)According to the idea that the ability of an index to discriminate the default state is stronger, the index weight is bigger, give each index a weight and a credit rating model for small enterprises is established. According to the standard that the credit rating is higher, the default loss rate is lower, divide the companies into different credit rating and the credit rating system for small enterprises is established.(3)Through the empirical analysis of a sample of 3111 small enterprises, a small enterprise credit rating index system consisting of 18 indicators, such as the quick ratio, long term assets suitable rate, is established. On this basis, calculate the small business credit scoring by the model of credit rating of small enterprises, divide 3111 small enterprises into 9 credit ratings according to the credit score. The final results show that the index system established in this paper can significantly distinguish the default state of small businesses, and the divided credit ratings meet the standard that the credit rating is higher, the default loss rate is lower.