信用风险度量与商业银行贷款定价研究
详细信息    本馆镜像全文|  推荐本文 |  |   获取CNKI官网全文
摘要
贷款业务是我国商业银行的核心业务,也是其盈利的主要来源之一。贷款定价的合理与否,不仅影响商业银行的收益水平,更会影响其资产质量、客户结构和市场竞争力。目前,在内外部激烈的市场环境下,加之缺乏对借款人信用风险的科学度量,商业银行制定的贷款价格与其承受的信用风险并不对等。由于这种不对等,使得借款人发生违约行为时,商业银行无法获得相应的风险补偿。当大量违约事件同时出现时,商业银行的生存将面临巨大风险。因此科学度量借款人的信用风险并将其合理的反映在贷款价格中,对于商业银行提高贷款定价能力、建立科学合理的贷款定价机制、适应利率市场化的金融环境具有重要的理论意义和切实可行的现实意义。
     任何企业都不能不受其他因素影响而孤立地制定价格,商业银行也不例外。本文在对信用风险度量和贷款定价的相关理论和方法进行分析的基础上,首先从商业银行贷款活动所涉及的宏观及行业层面、交易对手层面和银行内部经营层面对相关影响因素进行分析,并给出相应量化指标;接着从贷款合约角度分析贷款期限和贷款发放时间对贷款利率的影响并进行相应量化;最后对银企关系进行了界定并给出各类银企关系对贷款利率影响的量化标准。
     贷款的信用风险是商业银行在定价中最难把握的不确定因素,其度量包括对借款人违约概率和贷款违约损失率两方面。对借款人违约概率的确定上,本文在对借款人违约原因进行分析的基础上,首先对借款人债务保障现金流影响因素进行研究,并从定性和定量两方面设定贷款定价流程终止标准;接着根据借款人债务保障现金流分布形态和分布特征确定其分布函数的方法和步骤;最后设定了借款人违约的标准并对债务保障比率临界值进行了研究。
     对贷款违约损失率的确定上,本文首先给出在无其他担保方式下借款人保障现金流对贷款偿付率计算的方法,然后分别研究了第三方保证、抵押、质押三种担保方式下贷款违约损失率的度量。根据混合Copula模型建立了借款人和担保人的联合违约分布函数,并研究了在联合违约概率下担保人对贷款的偿付率;应用ARCH、GARCH和t-GARCH模型对抵押物价值波动性进行描述的基础上,借鉴VaR理论构建抵押物(质押物)风险价值模型,并根据抵押物(质押物)风险价值对债务的比率判定抵押物(质押物)是否可以设置担保。
     在体现贷款信用风险的前提下,商业银行在制定贷款价格时还要实现“具有竞争力、风险可控下收益最大化、覆盖经营成本实现预期盈利”的目标。为综合体现贷款信用风险量化与商业银行经营目标的结合,本文根据各目标下贷款定价所参照的信用风险标准不同分别建立了单目标贷款定价模型;然后根据多目标优化理论建立体现商业银行经营目标的贷款定价模型,并提出分层平衡思想求解贷款利率;最后利用Markov链对信用风险变化过程进行了描述,在此基础上实现了即期贷款利率的远期变换。
     在上述研究的基础上,本文第六章以某上市公司为例,研究在第三方担保下的贷款利率的确定过程。借助Matlab工具分别对借款人、担保方的债务保障现金流的分布函数以及两者间的联合违约分布函数进行了拟合,并对相关参数进行了估计和检验,在此基础上确定主体违约概率和联合违约概率,借助违约损失率的计算方法得到预期损失。在单目标贷款定价的基础上,根据分层平衡思想求解得到借款人的贷款利率区间。
Credit is one of the key businesses of commercial banks in China; it is alsoone of the major sources of their earnings. The reasonable price of loans will notonly promote the income of commercial banks and it also improves the quality ofits assets, the customer structure and its competitiveness. At present, under thefierce external compete and without scientific measurement of the borrower'scredit risk, when commercial banks make out the loan price, it can not cover thecredit risk the banks to face. Thus, when borrowers are default, the commercialbanks can not obtain the corresponding risk compensation. At the same time,when a large number of events of default occur, the commercial banks will facegreat survival risk. So measure the borrower's credit risk scientifically and let itbe a part of considerations of a loan pricing will be important theoretical andpractical significance for commercial banks to improve their loan pricing abilities,establish reasonable loan pricing methods and adapt to the market interestfinancial environment.
     Any one can’t make price without considerate other factor, besidescommercial banks. Based on the study of the related theory and method of thecredit risk measurement and loan pricing, we analyze the relevant factors whichaffect the loan activities in the macro and industry level, the counterparty leveland bank management level, and give the corresponding quantization standards;then we analyze the impact on loan interests that made by loan term and loanrelease time in the loan agreements and quantify them; finally we defines therelationship between banks and enterprises, and make out the effect on loaninterest made by all kinds of relationship between banks and enterprises, and givethe way to quantify them.
     When commercial banks make out the price of a loan, the credit risk of theloan is the most indeterminate factor and its measurement including calculate the probability of borrower default and the LGD of the loan. When determine theborrower’s default probability, this paper, based on the analysis of the borrower’sdefault reasons, first do some study on borrower’s debt ensuring cash flow andfrom quality and quantity two aspects we set the stop point for a loan. Thenaccording to the distribution pattern and distribution characters of the borrower’sdebt ensuring cash flow we give the way and steps to determine its distributionfunction. At last, we give the rules to judge whether the borrower is default andalso set the threshold value of the rules.
     When calculate the LGD of the loan, we first give the method for calculatingthe payment rate and default loss rate of debt without other guarantees. Then wediscuss the measurement of default loss under three conditions: first, the debtwith others guaranteed; second, and the debt with pledge; third the debt withhypothecation. By Multi-Copula theory, we develop the joint distributionfunction for borrower and guarantor. We use ARCH, GARCH and t-GARCHmodel to explain the fluctuations in the value of the collateral and reference toVaR theory we establish collateral value risk model. According to mortgage(pledge) value risk ratio of debt to determine the value of mortgage (pledge) canset the security.
     Having reflected the credit risk the loan pricing should cover the followingtargets: the price should be competitive, maximize the income on the risk undercontrol, cover the cost and achieve the desired profit. In order tocomprehensively reflect the combination of the loan credit risk and commercialbanks’ management goal, the paper analysis we made, we established signaltarget loan pricing models according to the different credit risk that target whichmentioned above should references. Then based on the multi-targets optimizationtheory we give the loan pricing model which reflect the commercial banksmanagement targets and use tiered balance pricing scheme to work out loanpricing for the model; Finally, we use Markov chain to introduce the changes incredit risk and based on these we realized the transformation of the current loaninterest to long-term loan interest.
     Based on these studies, in sixth chapter we have a listed company as anexample to show how to calculate the loan price with guarantee. In this example,with the help of Matlab we make out the distribution functions and their parameters for the debt ensuring cash flows of borrower itself and warrantor,then determine the borrower’s default probability and their joint defaultprobability. We can also get default loss rate through the method we mentioned inthis dissertation. Based on the Loan Pricing in a single target and tiered balancepricing we can get the ranges of loan price for borrower.
引文
[1]巴塞尔银行监管委员会.信用风险管理原则[S].北京:中国金融出版社,2002.
    [2] Basel Committee on Banking Supervision.International convergence ofcapital measurement and capital standards a revised framework [S/OL].NewYork:2006[2009-3-30]http://www.bis.org/publ/bcbsca.htm.
    [3] CHARLES W.CALOMIRIS,T.PORNROJNANGKOOL.Relationshipbanking and pricing of financial serbices[J].Journal of Financial ServicesResearch,2009,(35):189-224.
    [4] PANKAJ KUMAR MASKARA, DONALD J. MULLINEAUX. Smallfirm capital structure and the syndicated loan market[J]. Journal ofFinancial Services Research,2011,(39):55-70.
    [5] MARK.PYLES,DONALD JMULLINEAX.Constraints on loan salesand the price of liquidity[J].Journal of Financial Services Research,2008,(33):21-36.
    [6] BRYAN STANHOUSE, DUANE STOCK. Managing the risk of loanprepayments and the optimal structure of short term lending rates[J].Annals of Finance,2008,(4):197-215.
    [7] EVA CATARINEU-RABELL, PATRICIAL JACKSON, DIMITRIOSP.TSOMOCOS.Procyclicality and the new Basel Accord-bank’s choiceof loan rating system[J].Economic Theory,2005,(26):537-557.
    [8] DONALD P.,ADAM B.ASHCRAFT.Using loan rates to measure andregulate bank risk: fidings and an immodest proposal[J]. Journal ofFinancial Services Research,2003,(24):181-203.
    [9] JORION P. Risk: measuring the risk in Value at Risk[J]. FinancialAnalyst Journal,1996:45-47
    [10] CROUHY,MICHEL,DAN GALAI.A comparative analysis of currentcredit risk models[J].Journal of Banking&Finance,2000,(24):59-117.
    [11] ACHARYA, VIRAL V., BHARATH S. Does industry-wide distressaffect default firms?—Evidence from creditor recoveries[J].Journal ofFinancial Economics,2007,85(3):787-821.
    [12] LAITINEN E K.Predicting a corporate credit analyst’s estimate bylogistic and linear models[J].International Review of Financial Analysis,1999,(2):97-121.
    [13] PLUTO K,TASCHE D.Thinking coherently[J].Risk,2005,(8):72-79.
    [14] VERONICA B,FALKENHEIM M,POWELL A.On the use of portfoliorisk models and capital requirements in emerging markets the case ofargentine[J]. The World Bank Economic Review,2002,(16):187-212.
    [15]牛锡明.我国商业银行实行贷款定价之研究[J].金融研究,1997,(10):15-20.
    [16]盛军.科学的贷款定价方法——RAROC定价法[J].上海金融,2000,(7):48-49.
    [17]周四军.商业银行贷款定价方法研究[J].统计与决策,2002,(5):11-12.
    [18]陈丽霞,陈玉祥.商业银行贷款定价方法研究[J].工业技术经济,2002,(3):117-118.
    [19]殷仲民,李蕊.利率市场化条件下国内商业银行贷款定价模型[J].统计与决策,2006,(3):21-22.
    [20]石蓉,耿香娥.VAR方法在银行贷款定价中的应用[J].统计与决策,2002,(9):28-29.
    [21]陈燕玲.论利率市场化后的贷款定价模式选择[J].安徽大学学报,2002,26(3):82-85.
    [22]蒋东明.论国外贷款定价模式及其对我国商业银行的启示[J].北京理工大学学报,2004,6(5):83-85.
    [23]陈高翔.利率市场化条件下贷款定价的研究[J].价格理论与实践,2004,(10):51-52.
    [24]龙筱刚.商业银行贷款定价问题探讨[J].价格月刊,2009,(6):12-14.
    [25]刘新毅.利率市场化下国内商业银行贷款定价问题的研究[J].南方金融,2005,(6):26-28.
    [26]徐晶.西方贷款定价模型比较及其对我国商业银行的启示[J].广州市经济管理干部学院学报,2005,(4):1-4.
    [27]张国兴.浅谈我国商业银行的贷款定价方法[J].金融与经济,2005,(8):45-46.
    [28]曹清山,邹玉霞,王劲松.商业银行贷款定价策略和模型设计[J].金融论坛,2005,(2):33-37.
    [29]陈雪,杜玉兰,周淼.我国商业银行贷款定价模式的研究[J].商业研究,2005,(18):9-11.
    [30]汪妮.中国商业银行贷款定价模式比较及其现实选择[J].经济研究导刊,2010,(19):52-53.
    [31]蒲建平,余剑.贷款定价的期权分析及对中国银行信用风险思考[J].北方经贸,2001,(12):113-114.
    [32]简志宏,李楚霖.考虑违约风险的信贷资产定价[J].武汉理工大学学报,2001,23(4):71-73.
    [33]庄新田,黄小原.基于信息不对称的银行贷款定价策略分析[J].系统工程,2002,20(3):20-23.
    [34]王来星.我国商业银行收益与风险对应的定价模型构建[J].武汉金融,2003,(5):12-15.
    [35]毕明强.基于贡献度分析和客户关系的商业银行贷款定价方法研究[J].金融论坛,2004,(7):44-50.
    [36]单国霞.利率市场化下商业银行贷款定价探讨[J].技术经济与管理研究,2004,(4):103-104.
    [37]袁桂秋,金能.无违约风险的可调支付利率抵押贷款定价原理[J].系统工程学报,2004,19(2):135-140.
    [38]金雪军,毛捷.违约风险与贷款定价:一个基于期权方法和软预算约束的新模型[J].经济学,2007,6(4):1217-1238.
    [39]许玥.贷款定价的区域性策略比较[J].现代商业,2007,(17):36-37.
    [40]谭德俊,彭建刚.基于“三性”平衡的商业银行贷款定价[J].系统工程,2007,25(8):104-107.
    [41]戴立勇.商业银行贷款定价中的信用风险评价问题初探[J].西部金融,2007,(8):34-35.
    [42]于渤,郑画.基于客户贡献度的客户盈利贷款定价方法研究[J].技术经济与管理研究,2005,(3):98-99.
    [43]隋聪,迟国泰,闫达文.基于DEA二分法的贷款定价模型[J].预测,2009,(5):27-31.
    [44]于久洪,张剑.基于贝叶斯博弈的银行贷款定价研究[J].统计与决策,2010(7):138-140.
    [45]苏志明.我国商业银行贷款定价模型设计[J].技术经济与管理研究,2009(2):115-117.
    [46]黄春霞,何泽荣.基于Credit Metrics技术的RAROC贷款定价模型[J].经营管理者,2008,(15):5-7.
    [47]周凯.基于RAROC的商业银行贷款定价模式研究[J].世界经济与政治论坛,2008,(2):106-110.
    [48]陆学佳.基于RAROC的商业银行贷款定价机制研究[J].上海金融学院学报,2007,(5):39-44.
    [49]刘冬雨.我国商业银行贷款定价模式研究[J].商场现代化,2006,(28):365-366.
    [50]刘新军,周鸿卫.商业银行RAROC贷款定价方法的改进与应用[J].统计与决策,2009,(6):164-167.
    [51]杨继光,刘海龙,许友传.基于信用风险经济资本测度的的贷款定价研究[J].管理评论,2010,(7):33-38.
    [52]戴国强,吴许均.基于违约概率和违约损失率的贷款定价研究[J].国际金融研究,2005,(10):43-48.
    [53]周玮,杨兵兵,陈宏等.商业银行违约概率测算相关问题研究[J].国际金融研究,2005,(7):67-72.
    [54]赵跃琼,严广乐.KMV模型在银行贷款定价中的应用[J].上海理工大学学报,2006,28(1):23-26.
    [55]龙海明,邓太杏.基于信用评级的商业银行贷款风险定价模型[J].湖南大学学报,2007,34(2):88-92.
    [56]何晓光,许友传.贷款定价的信用风险成本的计量方法[J].金融与经济,2009,(2):72-74.
    [57]吴青.信用风险度量技术的最新发展及其在贷款定价中的应用[J].国际金融研究,2006,(9):15-23.
    [58]蒋明君.贷款理论定价与实际操作的差异及对策[J].现代金融,2007,(8):7-8.
    [59]殷鹏飞.内部评级技术的小企业贷款定价模型研究[J].经济研究导刊,2008,(16):23-25.
    [60]徐宝林,丁建平.贷款定价:从信用评级到经济资本[J].银行家,2006(3):94-99.
    [61]范永华.浅议商业银行贷款定价的路径选择[J].金融纵横,2008,(3):53-56.
    [62] MERTON R. On the pricing of corporate debt: the risk structure ofinterest rates[J].Journal of Finance,1974,(29):449-470.
    [63] LONGSTAFF F,SCHWARTA E.Valuing credit derivatives[J].Journalof Fixed Income,1995,(6):6-12.
    [64] DUFFIE D,SINGLETON K.Modeling term structures of default ablebonds[J].Review of Financial Studies,1999,(12):687-720.
    [65] CROUHY M,GALAI D,MARK R.A comparative analysis of currentcredit risk models[J].Journal of Banking&Finance,2000,(24):59-117.
    [66] KARATZAS,SHREVE S E.Brownian motion and stochastic calcu-lus[M].Springer,2nd edition,1991.
    [67] GIACOMO DE LAURENTIS. Corporate banker’s role and credit riskmanagement[M]. Strategy and Organization of Corporate Bank-ing.2005:107-137.
    [68] WILSON T.Portfolio credit risk,I.Risk.1987,10(9).
    [69] WILSON T.Portfolio credit risk,II.Risk.1987,10(10).
    [70] JARROW R A,LAND D,TURNBULL S M.A markov model for theterm structure of credit risk spreads[J].Review of Financial Studies,1997,(10):481-523.
    [71] BRODY D, HUGHSTON L, MACRINA A. Information-based assetpricing[J].International Fournal of Theoretical and Applied Finance,2008,(11):107-142.
    [72] BEAVER R. Financial ratios as predictors of failure [J]. Journal ofAccounting Research,1969,(4):71-111.
    [73] BEAVER R. Market prices, financial ratios and the prediction offailure[J].Journal of Accounting Research,1968,6(2):179-192.
    [74] ALTMAN E I.Financial ratios,discriminate analysis and the prediction ofcorporate bankruptcy[J].Journal of Finance,1968,(13):589-609.
    [75] ALTMAN E I,Hardeman R G,Narayanan P.ZETA analysis:a newmodel to identify bankruptcy risk of corporations[J].Journal of Bankingand Finance,1977,(1):29-54.
    [76] DEAKIN E B, A discriminate analysis of prediction of businessfailure[J].Journal of Accounting Research,1972,10(1):71-111.
    [77] OHLSON.Financial ratios and the probabilistic prediction of bankruptcy[J].Journal of Accounting Research,1980,(18):109-131.
    [78] CALONIRIS C, MASON J. Credit card securitization and regulatoryarbitrage[J].Journal of Financial Service Research,2004,(26):5-27.
    [79] HAIBIN ZHU.Unempirical comparison of credit spreads between thebond market and the credit default swap market[J].Journal of FinancialService Research,2006,(29):211-235.
    [80] MIHTON B A,STULTZ R,WILLIAMSON R.How much do banks usecredit derivative to reduce risk [D]. Working Paper, Ohio StateUniversity,2009:3-15.
    [81] PARLOUR C A,WINTON A.Laying off credit risk loan sale versuscredit default swaps[D]. Working Paper, University of California atBerkeley,2008:5-20.
    [82]约翰考埃特,爱德华爱特曼,保罗纳拉亚南.演进着的信用风险管理——金融领域面临的巨大挑战[M].石晓军,张振霞,译.北京:机械工业出版社,2004:45-54.
    [83] MESSIER W F,HANSEN J V.Inducing rule for expert system develop-ment:An example using default and bankruptcy data[J].Managementscience,1988,34(2):1403-1415.
    [84] ODOM M, SHARDA R. A neural network model for bankruptcyprediction[M]. CA: Int. Joint Conf. Neural Networks,1990:49-60.
    [85] Moody’s Investors Service Corporate.Moody’s quantitative risk’spublic firm risk model[R/OL].New York:2003[2007-9-20].http://www.Moodyssqra.com.
    [86] LEE K, HAN I, KWON Y. Hybrid neural network models forbankruptcy prediction[J].Decision Support System,1996,(18):63-72.
    [87] BACK B, LAITINEN T, ERE K. Neural networks and geneticalgorithms for bankruptcy prediction[J].Expert System Appl.,1996:407-413.
    [88] MARTINELLI E,CARVALHO A D,REZENDE S.Rules extractionsfrom banks’ bankrupt data using connectionist and symbolic learningalgorithms[M].New York:Computational Finance Conf.,1999:79-90.
    [89] YANG Z, PLATT M, PATUWO B. Artificial neural networks inbankruptcy prediction: general framework and cross-validationanalysis[J].European Journal of Oper.Res.,1999,(116):16-32.
    [90] TAM K. Neural network models and the prediction of bankbankruptcy[J].Omega,1991,(19):429-445.
    [91] TAM K,Kiang M.Managerial applications of the neural networks:thecase of bank failure predictions[J].Management Science,1992,(38):416-430.
    [92] SALCHENBERGER,CINAR E,LASH N.Neural networks:a new toolfor predicting thrift failures[J].Decision Sciences,1992,(23):899-916.
    [93] COATS P,FANT L.Recognizing financial distress patterns using a neuralnetwork tool[J].Financial Management,1993,(22):142-155.
    [94] ALTAMN E I,MARCO G,VARETTO F.Corporate distress diagnosis:comparisons using linear discriminate analysis neural networks[J].Journalof Banking and Finance,1994,(18):505-529.
    [95] SHIN K S,LEE T S,KIM H.An application of support vector machinesin bankruptcy prediction model[J]. Expert System with Application,2005,(28):127-135.
    [96] HILLEGEIST S A, KEATING E K, CRAM D P. Assessing theprobability of bankruptcy[J].Review of Accounting Studies,2004,9(1):5-34.
    [97]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,2002,(3):32-41.
    [98]李时春,周国祥.Credit MetricsTM和KMV模型在信用风险管理中的比较分析[J].农村经济与科技,2007,(8):66-69.
    [99]范南.Credit Metric模型及其对我国银行信用风险管理的借鉴[J].金融论坛,2002,(5):50-54.
    [100]梁琪,黄鹂皎.我国商业银行信贷风险管理体系构建探索[J].南开经济研究,2002,(6):61-65.
    [101]辛欣.基于Credit Risk+模型的科技型企业贷款组合研究[D].河南:河南大学硕士学位论文,2007:9-15.
    [102]张玲,杨贞柿,陈收.KMV模型在上市公司信用风险评价中的应用研究[J].系统工程,2004,22(11):84-89.
    [103]李磊宁,张凯.KMV模型的修正及在我国上市公司信用风险度量中应用[J].金融纵横,2007,(13):48-50.
    [104]赵保国.龙文征.信用评级中的违约率、违约概率研究[J].中央财经大学学报,2007,(1):38-43.
    [105]赵建卫.基于KMV的我国上市公司信用风险度量的实证研究[D].沈阳:东北财经大学硕士学位论文,2006:25-30.
    [106]孙燕妮.基于KMV模型的上市公司信用风险度量研究[D].大连:大连理工大学硕士学位论文,2007:19-21.
    [107]杨怀东,郭亚军,东明.分段效用决策的信用评级方法[J].系统工程,2004,22(4):86-90.
    [108]李家军.信用风险控制及其博弈分析[D].西安:西北工业大学博士学位论文,2005:36-45.
    [109]王春峰,万海晖,张维.组合预测在商业银行信用风险评估中的应用[J].管理工程学报,1999,(1):5-8.
    [110]陈弘.基于因子分析对信用评级方法的研究[D].长春:吉林大学硕士学位论文,2007:24-27.
    [111]周首华,杨济华,王平.论财务危机的预警分析——F分数模式[J].会计研究,1996,(8):8-11.
    [112]陈静.上市公司财务恶化预测的实证分析[J].会计研究,1999,(4):31-38.
    [113]陈瑜.对我国证券市场ST公司预测的实证研究[J].经济科学,2000,(6):57-67.
    [114]张玲.财务危机预警分析判别模型及其应用[J].预测,2000,(6):38-40.
    [115]张维,李玉霜,王春峰.递归分类树在信用风险分析中的应用[J].系统工程理论与实践,2000,(3):50-55.
    [116]卢声,任若恩,李清.中国上市公司财务困境模型的研究[J].北京航空航天大学学报,2001,14(1):22-26.
    [117]梁琪.企业信用风险的主成分判别模型及其实证研究[J].财经研究,2003,29(5):52-57.
    [118]张玲,曾维火.基于Z值模型的我国上市公司信用评级研究[J].财经研究,2004,30(6):5-13.
    [119]张玲.基于判别分析和期望违约率方法的信用风险度量及管理研究[D].长沙:湖南大学博士学位论文,2004:6-7.
    [120]于新花.企业信用风险评级模型的构建[D].青岛:中国海洋大学硕士学位论文,2004:6-10.
    [121]石晓军,任若恩,肖远文.边界Logistic违约率模型Bayes分析及实证研究[J].中国管理科学,2006,14(4):25-29.
    [122]伍韦,郭倩.基于判别模型的上市公司信用风险研究[J].财经界,2007,(7):86-87.
    [123]卢燕玲.基于Logistic模型的商业银行中等规模企业贷款信用风险研究[D].南京:南京理工大学硕士学位论文,2008:14-15.
    [124]王春峰,万海晖,张维.基于神经网络技术的商业银行信用风险评估[J].系统工程理论与实践,1999,(9):24-32.
    [125]傅荣,吴世农.我国上市公司经营失败风险的判定分析——BP神经网络模型和Fisher多类线性判定模型[J].东南学术,2002,(2):71-79.
    [126]韩静.基于概率神经网络的企业信用评级模型研究[J].工业技术经济,2005,24(9):41-43.
    [127]张鸿,丁以中.基于BP神经网络的企业信用评级模型[J].上海海事大学学报,2007,28(3):64-68.
    [128]高志. SOM神经网络模型在商业银行信用风险评估中的应用研究[J].电脑知识与技术.2007:498-499.
    [129]方先明,熊鹏,张谊浩.基于Hopfield神经网络的信用风险评价模型及其应用[J].中央财经大学学报,2007,(8):34-40.
    [130]何韧.银企关系与银行贷款定价的实证研究[J].财经论丛,2010,(1):57-63.
    [131] NELSON R B.An introduction to copulas[M].New York,Springer:1998.
    [132] ENGLE R F.Autoregressive heteroskedasticity with of the variance ofU.K.inflation[J].Econometrica,1982,(50):987-1008.
    [133] BOLLERSLEV T.Generalized autoregressive conditional heteroskedasti-city[J].Journal of Economics,1986,(31):307-327.
    [134] NELSEN R B. Dependence and order in families of ArchimedeanCopulas[J].Journal of Multivariate Analysis,1997,(60):111-122.