A Model Research For SME Guarantee Institution Evaluating SME Credit Rate

SME has become the important roles in the modern society. However, since the in advanced equipment and technology, being pressed for pledge, and so on, financing has become the difficulty for the SMEs. For sake of solving these difficulties, On base of research the foreign country's ways, our country has established the SME credit guarantee institutes. Gui zhou province began to set up SME credit guarantee institute in 1997. About 40 institutes have been set up. But because all kinds of difficulties, these institutes' development encounter the obstacle, scientific ways of setting up credit evaluation system are necessary for solving the obstacle that the SME credit guarantee institutes is encountering.In this article, first, I introduce the conventional credit evaluation methods, but the evaluation result of expert method depends on the experiences and abilities of assessors. The result is subjective and the objectivity and the just are hard to be guaranteed. Otherwise, statistic classification methods cannot solve the nonlinear highly related problems. The research purpose of this article to make up frothed is advantage of the conventional credit evaluation methods and to apply the powerful tool-BP Neural Network technique which studies complexity to credit evaluation of the enterprise so that more efficient practical credit evaluation methods can be achieved. This is the first time to use BP ANN in the researching realm of SME credit guarantee institutes.The data that I use in this article comes from the investigation from the 9th 2004 to the 5^th 2005 by the help of my teacher. I selected 137 enterprises from these enterprises being investigated. By means of the MATLAB's ANN toolbox, I trained and tested respective data. The rate of conection is 74.5 %. So as a new way of credit evaluation, this result proves that this model is effective and feasible.