Based On The Farmers Credit Rating Of The Logistic Regression Model

With the gradual advancement of global economic integration and the fast development of our country market economy, the industrial structure of the rural areas has undergone tremendous changes, more and more farmers apply for loans quickly developing farming and breeding industry, as farmers brought substantial revenue. However, the current rural financial system is half-baked and the unique credit risk farmers, resulting in the problems of the peasant credit rating credit arising in the course of the evaluation criteria is not yet standardized and farmers credit rating distortion, With the covariant risks of farmers as well as lower operating level of skills, more and more farmers break his word which has constraint credit institutions grant loans to farmers, and the majority of the assets of rural credit cooperatives are also shrinking so serious in recent years. How to distinguish the farmers whom the higher credit rating farmers (good customers) and the Lower credit rating farmers (bad customers) , then the credit institutions to achieve the sustained and healthy development and the farmer will solve the shortage of funds properly, which this is beneficial for all parties. Therefore, to establish an effective suitable Farmers credit rating criterion has the very important realistic meanings for the credit institutions and the farmer. The purpose of this paper is to analysis the various indicators, and found significantly higher credit rating distinction between the farmers (good customers) and the lower credit rating of the rural households (bad customers) indicators, and adopted the model of farmers return the loan got the probability of return, then get the credit scoring, finally get the farmers credit rating criteria which for the relevant parties to provide basis for decision-making and reference points.On the basis of forefathers' research results, using theoretical analysis and the analysis of the empirical analysis, on the basis of the analysis of credit risk uniqueness of the farmers drawn a conclusion: the farmers of credit ratings is very importance; And then analysis the problem of the current credit rating, drawn a conclusion: Objective credit ratings on the need to quantify are very important. Next this article's research object is the farmer who in order to development cultivation and aquaculture industry to apply for loans. We selected 160 farmers who belong to group of serious breaches (bad customers) and 250 farmers (good customers) who have never default as a training sample. In virtue of SPSS, using factor analysis of the assets of farmers operating indicators,debt indicators,and other indicators of moral prestige and so on get five key indicators, next utilizes the logistic regression analysis to select four indicators which have the remarkable contribution on the model forecast, and the four variables as the initial variable to establish farmers' credit default model. Next uses 410 training samples and 115 examining simples to test the validity of models. The results showed that: the overall contractor for the 85.4 percent, and the predictive rate of accuracy of good customers is to be itself is 80 per cent, the predictive rate of accuracy of bad customers is to be itself is 93.8 percent; the predictive rate of accuracy is 85.4 percent, and the predictive rate of accuracy of good customers is to be itself is 85.26 per cent, the predictive rate of accuracy of bad customers is to be itself is 80 percent. Obviously, the model gets a good predictive effect and the certain application value. Finally, use the cluster analysis methods of classification get the farmer credit rating criteria.In this paper, an analysis of credit to provide the credit audit staff, drawn such conclusions: per capita income of last year, per capita total expenditure last year, the per capita possession of agriculture and forestry specialty prestige of the four indicators, as well as a good reflection of the credit status of farmers, which have good ability to predict the probability of future, in the course of the farmers credit exist more Strong signal indicators; credit institutions should accept more affluent level of prestige but good borrower, refused to those who do not poverty but with a relatively low level of borrowing prestige; per capita expenditure reflect the default status of rural households in a certain extent; for farmers planting size, the scale doesn't mean that the low default, it should be with reference to its technical quality of labor, farmers should be planting size and quality of its labor force in line technology.