| 1,498 | 84 | 58 |
| 下载次数 | 被引频次 | 阅读次数 |
本文首次构建了基于非参数随机森林(Random Forest,RF)的住房贷款违约风险评估模型,利用某大型银行个人住房贷款数据,研究了借款人特征、贷款特征、房产特征和经济文化特征等因素对贷款违约的影响。实证研究发现已偿还比例、利率、贷款收入比、额度等是贷款违约最重要的影响因素,并且RF方法的预测准确率明显高于logistic模型等其他方法。此外,本文还研究了利率调整对贷款违约的影响,发现利率对违约率的影响是负方向的,且呈不对称性和非线性。
Abstract：This paper proposed a housing mortgage default risk forecasting model based on non-parametric random forest at first. Then by using the housing mortgage database from a big famous bank in China,this paper studied the effect of housing mortgage default according to borrowers' characteristics,loan characteristics,housing characteristics and local economic and cultural characteristics. The empirical study found that the proportion which had been repaid,interest rate,ratio of loan to income,loan amount were the most important factors. The results also showed the prediction accuracy of RF were much higher than other methods such as logistic regression. In addition,this paper also studied how the interest rate affected mortgage default,finding that interest rate had negative effect,which were asymmetry and nonlinear,on the mortgage default.
[1]Chen T.H.and Chen C.W.Application of data mining to the spatial heterogeneity of foreclosed mortgages[J].Expert Systems with Applications,2010,37(2):993-997.
[2]Medema L.,Koning R.H.,and Lensink R..A practical approach to validating a PD model[J].Journal of banking and finance,2009,33(4):701-708.
[3]Bellotti,T.and Crook,J.Support vector machines for credit scoring and discovery of signicant features[J].Expert Systems with Applications,2009,36(2):3302-3308.
[4]Lee Y.C.Application of support vector machines to corporate credit rating prediction[J].Expert Systems with Applications,2007,33:67-74.
[5]Twala,B.Multiple classifier application to credit risk assessment[J].Expert Systems with Applications,2010,37(4):3326-3336.
[6]Bhekisipho T.Multiple classifier application to credit risk assessment[J].Expert Systems with Applications,2010,37(4):3326-3336.
[7]马宇.我国个人住房抵押贷款违约风险影响因素的实证研究[J].统计研究,2009(5):100-107.
[8]颜新秀.个贷违约率与宏观经济指标相关性研究[J].国际金融研究,2009(10):59-67.
[9]平新乔,杨慕云.信贷市场信息不对称的实证研究——来自中国国有商业银行的证据[J].金融研究,2009(3):1-18.
[10]Lawrenc E.C.and Arshadi N.A multinomial logit analysis of problem loan resolution choices in banking[J].Journal of Money,Credit and Banking.1995,27(1):202-216.
[11]Cowan,A.M.and Cowan,C.D.Default correlation:An empirical investigation of a subprime lender[J].Journal of Banking and Finance,2004,28(4):753-771.
[12]Qi M.and Yang X.L.Loss given default of high loan-to-value residential mortgages[J].Journal of Banking and Finance,2009,33(5):788-799.
[13]Gau G..W.A taxonomic model for the risk-rating of residential mortgages[J].The journal of business.1978,51(4):687-706.
[14]Magri,S.and Pico,R.The rise of risk-based pricing of mortgage interest rates in Italy[J].Journal of banking and finance,2011,35(5):1277-1290.
[15]Daglish T.What motivates a subprime borrower to default?[J].Journal of Banking and Finance,2009,33(4):681-693.
[16]方匡南,吴见彬,朱建平,等.信贷信息不对称下的信用卡信用风险研究——基于非参数随机森林模型的实证分析[J].经济研究,2010(1):97-107.
基本信息:
DOI：10.19343/j.cnki.11-1302/c.2013.10.008
中图分类号:F832.479
引用信息:
[1]方匡南,吴见彬.个人住房贷款违约预测与利率政策模拟[J].统计研究,2013,30(10):54-60.DOI:10.19343/j.cnki.11-1302/c.2013.10.008.
2013-10-15
2013-10-15