基于修正Benford律的财务危机预警Logistic模型及其应用  

Financial Early Warning Logistic Model Based on Corrected Benford’s Law and Its Application

作　　者：杨贵军 孙玲莉 周亚梦 石玉慧 YANG Gui-jun;SUN Ling-li;ZHOU Ya-meng;SHI Yu-hui(School of Statistics&China Center of Economics and Statistics Research,Tianjin University of Finance and Economics,Tianjin 300222,China;Shenzhen Institute of Advanced Technology,Chinese Academy of Sciences,Shenzhen 518055,China;Shenzhen Real Estate and Urban Construction Development Research Center,Shenzhen 518000,China;Digital banking management department,Bank of Nanjing,Nanjing 210005,China)

机构地区：[1]天津财经大学统计学院和中国经济统计研究中心,天津300222 [2]中国科学院深圳先进技术研究院,广东深圳518055 [3]深圳市房地产和城市建设发展研究中心,广东深圳518000 [4]数字银行管理部南京银行总行,江苏南京210005

出　　处：《数理统计与管理》2021年第4期585-595,共11页Journal of Applied Statistics and Management

基　　金：国家社会科学基金项目(20CTJ009,20ATJ008);国家自然科学基金资助项目(11471239);2019年度天津市哲学社会科学规划课题重点课题(TJTJ19-001)。

摘　　要：财务信息失真降低了财务危机预警模型的预测正确率。针对财务指标数据为有界子集的情形,本文引入基于修正Benford因子的财务危机预警Logistic模型,包含代表财务数据质量的修正Benford因子,改善财务危机预警Logistic模型的预测效果。关于中国上市公司的数据分析发现,修正Benford因子可以提高财务危机预警Logistic模型的预测正确率,降低第一类误判率和第二类误判率。Financial information distortion reduces the predictive accuracy of the financial early warning model.When financial indicator data is a bounded subset,this article introduces financial early warning Logistic model based on corrected Benford’s law which contains the corrected Benford’s factors to represent the quality of financial data.Based on the Chinese listed companies used as samples,the obtained results show that Benford’s factors improve the predictive accuracy of financial early warning model and reduce type-I error and type-II error.

关 键 词：财务危机预警 财务信息失真 修正Benford因子 预测准确性 

分 类 号：F830.9[经济管理—金融学] O211.3[理学—概率论与数理统计]