我国农林牧渔业短期贷款企业违约判别实证分析  

An Empirical Analysis of Default Judgment of Enterprise on Short-Term Loans in Agriculture, Forestry, Husbandry and Fishery Industries

作　　者：管七海 

机构地区：[1]招商银行博士后科研工作站

出　　处：《金融论坛》2006年第1期14-19,共6页Finance Forum

基　　金：国家自然科学基金项目(项目号No.70171005);国家十五攻关项目(项目号No.2001BA102A06-07-01)。

摘　　要：近几年,我国农林牧渔业短期贷款企业的违约严重程度一直居所有行业之首,从跨行业的角度评估该行业短期贷款企业的违约具有重要意义。本文基于全国跨银行的贷款企业海量数据库样本,针对农林牧渔业的短期贷款企业进行了分规模和分地区样本的多元判别分析模型、Logistic模型与神经网络模型等的构建与实证探索,进而找出了影响我国农林牧渔业企业违约的关键变量,构建了最佳违约判别模型。这些关键变量和判别模型对中国人民银行和各商业银行监测该行业企业的信用风险具有重要的参考价值。In recent years, the seriousness of defauh by enterprises on short-term loans in agriculture, forestry, husbandry and fishery industries tops that of all industries. So it is of great importance to evaluate from a trans-industry perspective the breaching acts of enterprises on short-term loans in the industry. Based on numerous data base samples of enterprises on nationwide inter-bank loans, this paper establishes separately a pluralistic judgment model, a logistic model and a neurological network model for such enterprises in light of their size and regions. Through an empirical study, key variables affecting enterprises on default in the industry are found and best default judgment model is established. These variables and models are believed to referentially significant to People＇s Bank of China and commercial banks in monitoring credit exposures of enterprises in the sector.

关 键 词：农林牧渔业 短期贷款企业 违约模型 多元判别模型 LOGISTIC模型 神经网络模型 

分 类 号：F832.4[经济管理—金融学]