家庭农牧场信用评价指标体系构建方法与应用——基于Wald检验-相关分析

The Construction Method and Application of Credit Evaluation Indicator System of Family Farm and Pasture:Based on Wald Test-Correlation Analysis

现有对家庭农牧场的研究大多是从新型农业经营主体融资角度或从某一方面对新型农业经营主体的影响进行模型分析或定性分析,很少有针对建立家庭农牧场信用评价指标体系方面的研究,而这恰是解决家庭农牧场融资困境的有效办法之一。科学合理的信用评价指标体系可以促使金融机构开发家庭农牧场信贷产品,进而推动家庭农牧场快速、长期、稳定发展。通过Logistic回归中的Wald显著性判别和皮尔逊相关分析两种方法对家庭农牧场的信用评价指标进行两次定量筛选,保证最终构建的包含成立日期、经营土地总块数等11个指标在内的家庭农牧场信用评价指标体系既具有不同信用区别能力又具有指标信息不重复性。内蒙古12个盟市的共722家家庭农牧场ROC曲线表明:AUC=0.896>0.70,说明利用Wald检验与相关分析法相结合所构建的家庭农牧场信用评价指标体系是有效的。

Most of the existing researches on family farm and pasture are model analysis or qualitative analysis on the impact of new agricultural management entities from the financing perspective of the new type of agricultural management entities or from a certain aspect;There are few researches on the establishment of credit evaluation indicator system for family farm and pasture.And this is one of the key ways to solve the financing dilemma of family farm and pasture.Scientific and reasonable credit evaluation indicator system can promote financial institutions to develop family farm and pasture credit products,and then promote the rapid,long-term and stable development of family farm and pasture.Two methods of Wald significance discrimination and Pearson correlation analysis in Logistic regression were used to quantitatively screen the credit evaluation indicators of family farms and pastures.The credit evaluation indicator system of family farm and pasture that is finally constructed,including 11indicator such as the date of establishment and the total number of plots of managed land,has not only different credit distinction ability but also non-repeatability of indicator information.The ROC curve of 722 family farms and pastures in 12 cities of Inner Mongolia showed that AUC=0.8960.70,indicating that the credit evaluation indicator system of family farms and pastures constructed by Wald test and correlation analysis method is effective.