我国民营上市公司信用风险度量研究
|
摘要
随着我国经济的快速发展,民营上市公司通过兼并、合资及独资等投资形式,加速在多个经济领域的扩张。为弥补其因扩张而日益彰显的资金不足问题,民营上市公司纷纷拓展其融资渠道。随着近几年我国债券市场的日益规范,债券融资渐渐受到民营上市公司的热捧,如短期融资券、公司债券的发行等。然而,与此相对应的是我国民营上市公司被特别处理的家数也呈逐年增加的趋势,这反映了一些民营上市公司的内部管理混乱、治理结构失衡、投资决策出现严重失误、信用管理出现危机等问题,从而使民营上市公司面临退市风险的考验,债权人、投资者的潜在利益也受到重大威胁。为了能及早预测到负债企业在近年内的信用状况,以期能给我国民营上市公司目前存在的信用风险的识别和控制提供一些有益的帮助,本文利用主成因子分析、Logistic回归模型分析,从定量和定性两个角度对我国民营上市公司信用风险进行实证研究。
本文首先根据证监会制定的关于我国上市公司被特别处理的原则,选取了45家民营上市公司作为经营失败企业样本,并按照同规模、同行业、同年份等原则按比例选取了90家非ST民营上市公司作为经营正常企业样本,并将全部样本分为估计样本和测试样本。其次,根据影响我国上市公司短期信用风险的若干因素,并结合指标变量的选取原则,选取了六大类二十五种企业财务比率作为样本变量,最后,利用主成分分析和Logistic回归模型分析对估计样本建模,并利用测试样本检验模型的判别准确率。
检验结果显示,模型的预测准确率达到88.3%以上。根据分析结果,从中得到了几点关于我国民营上市公司信用风险预测的结论,如主成分因子分析能够很好地解决Logistic回归模型问题,预测民营上市公司信用风险要坚持定量分析和定性分析相结合,越接近被特别处理的日期,预测结果的可靠性越高等。
With the fast development of economy in china, the Private listed companies, through each kind of investment form, such as annexation, joint venture, single proprietorship, expand their economic domain acceleratly. For making up the fund shortage caused by expantion, the Private listed companies expand their financing channels one after another.In recent years, with the standardization of band market in china; the Private listed companies gradually high on bond financing, such as the issuing of finance bond and company bond. However, correspondingly, the amount of Private listed companies special treated by China Security Regulatory Commission increases year by year. This reflects some problems of Private listed companies, such as confusion of inner management, unbalance of governance structure and significant lapsus of capital investment decision, Credit Management crisis and other issues ,as a result, the Private listed companies faces the risk of "withdrawing market", also the potential interest of the creditor and investor faces enormous losses. In order to forcast the credit condition of bond-issuing company in two or three years earlier, so providing some helpful help for the Private listed companies in the recognition and controlment of credit risk, the paper uses principal component logistic regression, from quantitative analysis and qualitative analysis, to carry on empirical research in credit risk.
At first, according to the principle of ST, the paper chooses 45 ST companies as bad-manage sample, and according to the principle of same scale, same industry and same period, chooses 90 no ST companies as well-manage sample, besides, the paper divides all samples into two kinds: estimate sample and forcast sample. Secondly, according to the factors affect the credit risks of Private listed companies, combined with the choosing principle of index variables. At last, the paper builds principal component logistic regression model, and uses forcast samples to inspectorder of accuracy.
Test result shows, the forcasting exact rate model reaches 88.3percent. According to the analysis result, the author propose several conclusions, such as, the principle component analysis can resolve the problem of multicollinearity in the logit regression model; the quantitative analysis and the qualitative analysis should be combined when analyse credit risk; the closer to the period of ST, the higher of forcast predict rate, and so on.
本文首先根据证监会制定的关于我国上市公司被特别处理的原则,选取了45家民营上市公司作为经营失败企业样本,并按照同规模、同行业、同年份等原则按比例选取了90家非ST民营上市公司作为经营正常企业样本,并将全部样本分为估计样本和测试样本。其次,根据影响我国上市公司短期信用风险的若干因素,并结合指标变量的选取原则,选取了六大类二十五种企业财务比率作为样本变量,最后,利用主成分分析和Logistic回归模型分析对估计样本建模,并利用测试样本检验模型的判别准确率。
检验结果显示,模型的预测准确率达到88.3%以上。根据分析结果,从中得到了几点关于我国民营上市公司信用风险预测的结论,如主成分因子分析能够很好地解决Logistic回归模型问题,预测民营上市公司信用风险要坚持定量分析和定性分析相结合,越接近被特别处理的日期,预测结果的可靠性越高等。
With the fast development of economy in china, the Private listed companies, through each kind of investment form, such as annexation, joint venture, single proprietorship, expand their economic domain acceleratly. For making up the fund shortage caused by expantion, the Private listed companies expand their financing channels one after another.In recent years, with the standardization of band market in china; the Private listed companies gradually high on bond financing, such as the issuing of finance bond and company bond. However, correspondingly, the amount of Private listed companies special treated by China Security Regulatory Commission increases year by year. This reflects some problems of Private listed companies, such as confusion of inner management, unbalance of governance structure and significant lapsus of capital investment decision, Credit Management crisis and other issues ,as a result, the Private listed companies faces the risk of "withdrawing market", also the potential interest of the creditor and investor faces enormous losses. In order to forcast the credit condition of bond-issuing company in two or three years earlier, so providing some helpful help for the Private listed companies in the recognition and controlment of credit risk, the paper uses principal component logistic regression, from quantitative analysis and qualitative analysis, to carry on empirical research in credit risk.
At first, according to the principle of ST, the paper chooses 45 ST companies as bad-manage sample, and according to the principle of same scale, same industry and same period, chooses 90 no ST companies as well-manage sample, besides, the paper divides all samples into two kinds: estimate sample and forcast sample. Secondly, according to the factors affect the credit risks of Private listed companies, combined with the choosing principle of index variables. At last, the paper builds principal component logistic regression model, and uses forcast samples to inspectorder of accuracy.
Test result shows, the forcasting exact rate model reaches 88.3percent. According to the analysis result, the author propose several conclusions, such as, the principle component analysis can resolve the problem of multicollinearity in the logit regression model; the quantitative analysis and the qualitative analysis should be combined when analyse credit risk; the closer to the period of ST, the higher of forcast predict rate, and so on.
引文
[1] W.Beaver.Financial ratios as predic:ors of failure[J]Journal of Accounting Research, 2006,(supplement): 71 — 102
[2] Altaian.Edward I. Financial ratios, Discriminant Analysis and the Prediction of Corporate Bankruptcy[J]. Journal of Finance, 2005,(23): 589-609
[3] Altman.Edward L, Haldeman.R.Cz, Narayanan. P. ZETAAnalysis: A New Model to Identify Bankruptcy Risk of Corporations[J]. Journal of Banking and Finance, 2007,(1): 29-54.
[4] Martin. D. Early Warning of Bank Failure: A logit Regression Approach[J]. Journal of Banking and Finance, 2007, (2): 249-276.
[5] Ohlson J.Financial Rations and the Probabilistic Prediction of Bankruptcy[J]. Accounting Research,1980,18(1): 109-130.
[6] Press S.J., Wilson S. Choosing between Logistic regression and discrimininant analysis[J]. J.Statist. Assoc, 2008, (73): 699-705.
[7] Eisenbeis R. A.. Pitfalls in the application of discriminant analysis in business[J].Journal of Finance, 1977,(32): 87-90.
[8] Coats. P., Fant. L.. Recognizing Financial Distress Patters Using Neural Network Tool[J].Financial Management, 1993,(3): 142-155.
[9] Odour M.D., Sharda R. A neural network for bankruptcy prediction[J]. International Joint Conference on Neural Network, 2007,(2): 163 — 168.
[10] Altaian. Edward L, Macro, A.. Corporate distress diagnosis: comparisons using lineardiscriminant analysis and neural networks (the Italian experience)[J]. Journal of Banking and Finance,2004,(18): 505-529.
[11] J.H. Min, Y.C. Lee. Bankruptcy prediction using support vector machine with optimal choice of kernel function parameters[J]. Expert Systems with Applications,2007,(28): 603-614
[12] Henley. W.E., and D.J. Hand.A k-nearest-neighbor classification for assessing consumer credit risk[J].The Statistician, 2005,45 (1): 77-95
[13] Lundy M 二 Cluster Analysis in Credit; Coring. Credit Scoring and Credit Control[M]. New York:Oxford University Press, 2006: 33-52
[14]Tang.K.Y,Kiang.M..Managerial Applications of Neural Networks:The Case of Bank Failure Predictions[J].Management Science,2008,(38):927-947.
[15]卢世春,欧阳植.商业银行信用风险跟踪预警监测模型.数量经济技术经济研究[J].2008,(1):59-61
[16]陈静.上市公司财务恶化预测的实证分析[J].会计研究.2007,(4):31-38
[17]张玲.财务危机预警分析判别模型及应用[J].预测.2007,(6):38-40
[18]陈晓,陈治鸿.中国上市公司的财务困境预测[J].中国会计与财务研究.2007,(9):55-72.
[19]吴世农,卢贤义.我国上市公司财务困境的预测模型研究.经济研究[J].2008,(6):46-55
[20]张后奇,刘月平等.上市公司财务危机预警系统:理论研究和实证分析.长城证券课题组.上海证券交易所课题.2008:133-141
[21]沈沛龙,任若恩.现代信用风险管理模型和方法的比较研究[J].经济科学2005,(3):32-41
[22]朱小宗,张宗益等.现代信用风险度量模型剖析与综合比较分析[J].财经研究.2007,(9)33-41
[23]李秉祥.基于期望违约率模型的上市公司财务困境预警研究[J].中国管理科学.2006,(5):12-16
[24]约翰.B.考埃特爱德华.t.埃特曼保罗.纳拉业南—石晓军、张振霞译.演进着的信用风险管理:金融领域面临的巨大挑[M].北京:机械土业出版社.2006:89-93
[25]彼得·S}罗斯.商业银行管理[M].北京.经济科学出版社,2007:26-42
[26]约翰.B.考埃特著,石晓军,张振霞译.演进着的信用风险管理[M].机械工业出版社,2007:286-287
[27]王济川、郭志刚,Logistic回归模型一方法与应用,北京:高等教育出版杜,2007
[28]Beaver,W.H.Financial ratios as predictors of failure[J].Journal of Accounting Research,2006,(18):71-111
[29]李晓著,我国上市公司财务危机预警及其管理对策研究,[硕士学位论文],北京:北京交通大学,2006:178-184
[30]Gentry,J.A.,Newbold,P.&Whitford,D.T.Classing bankruptcy firms with funds flow components financial rations,and bankruptcy[J].Journal of Accounting Research,1985,23(1):146-160
[31]傅强,李永涛,基于logistic模型的上市公司信用风险评价,华东经济管理,2007,(9):95-97
[32]潘玉叶,台湾股票上市公司财务危机预警分析,[博士学位论文],浙江大学,2005:39
[33]石晓军,肖远文和任若恩,Logistc违约率模型的最优样木配比和分界点研究,财经研究,2006,31(9):38-48
[34]Matthew Kurbat&Irina Korablev,"Methodology far Testing the Level of the EDF(tm)Credit Measure".2003:128-133
[35]马庆国,管理统计:数据获取统计原理、SPSS工具与应用研究,北京:科学出版社,2008:315-335
[36]王福林,个人住房抵押贷款违约风险影响因素实证研究,[博士学位论文],杭州:浙江大学,2008 159-162
[37]江苏省企业信用管理协会,上海万事达商业征信服务有限公司,企业信用管理操作实务,北京:中国方正出版社,2006:342-246
[38]陆纹岱,SPSS for Windows统计分析,北京,电子工业出版社,2007:199-203
[39]Andreas Charitou,Lenos Trigeorgis."Option-Based Prediction".University of Cyprus Working Paper Social Science Research Network Electronic Paper Collection,2007:321-327
[40]Theophanous,T.P.Corporate failure prediction models for the U.S.manufacturing and retailing sectors[D].PhD.Dissertation of University of New York.University Microfilms.Tntemationals,2005:277-289
[2] Altaian.Edward I. Financial ratios, Discriminant Analysis and the Prediction of Corporate Bankruptcy[J]. Journal of Finance, 2005,(23): 589-609
[3] Altman.Edward L, Haldeman.R.Cz, Narayanan. P. ZETAAnalysis: A New Model to Identify Bankruptcy Risk of Corporations[J]. Journal of Banking and Finance, 2007,(1): 29-54.
[4] Martin. D. Early Warning of Bank Failure: A logit Regression Approach[J]. Journal of Banking and Finance, 2007, (2): 249-276.
[5] Ohlson J.Financial Rations and the Probabilistic Prediction of Bankruptcy[J]. Accounting Research,1980,18(1): 109-130.
[6] Press S.J., Wilson S. Choosing between Logistic regression and discrimininant analysis[J]. J.Statist. Assoc, 2008, (73): 699-705.
[7] Eisenbeis R. A.. Pitfalls in the application of discriminant analysis in business[J].Journal of Finance, 1977,(32): 87-90.
[8] Coats. P., Fant. L.. Recognizing Financial Distress Patters Using Neural Network Tool[J].Financial Management, 1993,(3): 142-155.
[9] Odour M.D., Sharda R. A neural network for bankruptcy prediction[J]. International Joint Conference on Neural Network, 2007,(2): 163 — 168.
[10] Altaian. Edward L, Macro, A.. Corporate distress diagnosis: comparisons using lineardiscriminant analysis and neural networks (the Italian experience)[J]. Journal of Banking and Finance,2004,(18): 505-529.
[11] J.H. Min, Y.C. Lee. Bankruptcy prediction using support vector machine with optimal choice of kernel function parameters[J]. Expert Systems with Applications,2007,(28): 603-614
[12] Henley. W.E., and D.J. Hand.A k-nearest-neighbor classification for assessing consumer credit risk[J].The Statistician, 2005,45 (1): 77-95
[13] Lundy M 二 Cluster Analysis in Credit; Coring. Credit Scoring and Credit Control[M]. New York:Oxford University Press, 2006: 33-52
[14]Tang.K.Y,Kiang.M..Managerial Applications of Neural Networks:The Case of Bank Failure Predictions[J].Management Science,2008,(38):927-947.
[15]卢世春,欧阳植.商业银行信用风险跟踪预警监测模型.数量经济技术经济研究[J].2008,(1):59-61
[16]陈静.上市公司财务恶化预测的实证分析[J].会计研究.2007,(4):31-38
[17]张玲.财务危机预警分析判别模型及应用[J].预测.2007,(6):38-40
[18]陈晓,陈治鸿.中国上市公司的财务困境预测[J].中国会计与财务研究.2007,(9):55-72.
[19]吴世农,卢贤义.我国上市公司财务困境的预测模型研究.经济研究[J].2008,(6):46-55
[20]张后奇,刘月平等.上市公司财务危机预警系统:理论研究和实证分析.长城证券课题组.上海证券交易所课题.2008:133-141
[21]沈沛龙,任若恩.现代信用风险管理模型和方法的比较研究[J].经济科学2005,(3):32-41
[22]朱小宗,张宗益等.现代信用风险度量模型剖析与综合比较分析[J].财经研究.2007,(9)33-41
[23]李秉祥.基于期望违约率模型的上市公司财务困境预警研究[J].中国管理科学.2006,(5):12-16
[24]约翰.B.考埃特爱德华.t.埃特曼保罗.纳拉业南—石晓军、张振霞译.演进着的信用风险管理:金融领域面临的巨大挑[M].北京:机械土业出版社.2006:89-93
[25]彼得·S}罗斯.商业银行管理[M].北京.经济科学出版社,2007:26-42
[26]约翰.B.考埃特著,石晓军,张振霞译.演进着的信用风险管理[M].机械工业出版社,2007:286-287
[27]王济川、郭志刚,Logistic回归模型一方法与应用,北京:高等教育出版杜,2007
[28]Beaver,W.H.Financial ratios as predictors of failure[J].Journal of Accounting Research,2006,(18):71-111
[29]李晓著,我国上市公司财务危机预警及其管理对策研究,[硕士学位论文],北京:北京交通大学,2006:178-184
[30]Gentry,J.A.,Newbold,P.&Whitford,D.T.Classing bankruptcy firms with funds flow components financial rations,and bankruptcy[J].Journal of Accounting Research,1985,23(1):146-160
[31]傅强,李永涛,基于logistic模型的上市公司信用风险评价,华东经济管理,2007,(9):95-97
[32]潘玉叶,台湾股票上市公司财务危机预警分析,[博士学位论文],浙江大学,2005:39
[33]石晓军,肖远文和任若恩,Logistc违约率模型的最优样木配比和分界点研究,财经研究,2006,31(9):38-48
[34]Matthew Kurbat&Irina Korablev,"Methodology far Testing the Level of the EDF(tm)Credit Measure".2003:128-133
[35]马庆国,管理统计:数据获取统计原理、SPSS工具与应用研究,北京:科学出版社,2008:315-335
[36]王福林,个人住房抵押贷款违约风险影响因素实证研究,[博士学位论文],杭州:浙江大学,2008 159-162
[37]江苏省企业信用管理协会,上海万事达商业征信服务有限公司,企业信用管理操作实务,北京:中国方正出版社,2006:342-246
[38]陆纹岱,SPSS for Windows统计分析,北京,电子工业出版社,2007:199-203
[39]Andreas Charitou,Lenos Trigeorgis."Option-Based Prediction".University of Cyprus Working Paper Social Science Research Network Electronic Paper Collection,2007:321-327
[40]Theophanous,T.P.Corporate failure prediction models for the U.S.manufacturing and retailing sectors[D].PhD.Dissertation of University of New York.University Microfilms.Tntemationals,2005:277-289