基于零膨胀模型的我国银企关系规模影响因素研究  

Empirical Determinant of Bank Firm Relationship Number—Study Based on Zero Inflated Models

作　　者：罗付岩[1,2] 赵佳星[2] LUO Fu-yan ZHAO Jia-xing(Modern Enterprise Management Research Center, Guilin University of Technology, Guangxi Guilin 541004, China School of Management, Guilin University of Technology, Guangxi Guilin 541004, China)

机构地区：[1]桂林理工大学现代企业管理研究中心,广西桂林541004 [2]桂林理工大学管理学院,广西桂林541004

出　　处：《数理统计与管理》2017年第2期351-360,共10页Journal of Applied Statistics and Management

基　　金：国家自然科学基金(71562007);广西高校人文社会科学重点研究基地基金资助(15YB004)

摘　　要：银企关系是学术界和实务界关注的焦点之一,然而,国内学者鲜有探讨银企关系数量的影响因素。本文使用我国A股上市公司2006-2013年的银企关系计数资料,利用零膨胀模型对企业建立银企关系规模的影响因素进行了分析。研究发现:规模大、资产负债率高、获利能力强的公司倾向于建立更多的银企关系;企业的长期负债率、第一大股东持股比例,是否是国有产权属性和企业的经营风险与银企关系的规模(数量)显著负相关;信贷合约的期限和信贷金额与银企关系的数量显著正相关;进一步比较了零膨胀模型与Poisson回归、负二项分布回归模型等计数模型,统计检验显示,零膨胀模型比较适合零值过多和过度离散的数据结构资料。Bank firm relationship becomes one of the focus in academics and practitioners.However,less domestic scholars explore the determinant of bank firm relationship number.This paper analyze the determinant of bank firm relationship number using A share 2006-2013 count data of listed company in China and using zero inflated model.The results show that the firms with Larger-scale,higher assets and liabilities rate,stronger profitability prone to multiple bank relationship,the relationship between longterm corporate debt ratio,the hold share proportion of the largest shareholder,state ownership enterprise,corporate operating risk and the bank firm relationship number is a significant negative correlation.The relationship between duration and the amount of credit contract and the bank firm number is significant positive.Further more,we compare the zero-inflated Poisson regression model and negative binomial regression models etc.count model,statistical tests show zero inflation model is suitable for a too much zero value and over dispersion data structure.

关 键 词：银企关系 计数资料 零膨胀模型 Poisson模型 

分 类 号：F275.5[经济管理—企业管理] O212[经济管理—国民经济]