住房抵押贷款定价模型研究综述  

Survey of residential mortgage pricing models research

作　　者：曹国华[1] 牟生洪[1] 

机构地区：[1]重庆大学经济与工商管理学院,重庆400044

出　　处：《系统工程学报》2006年第3期305-312,共8页Journal of Systems Engineering

基　　金：国家自然科学基金资助项目(70571089)

摘　　要：本文从研究工具和方法论入手,简述了国外研究住房抵押贷款定价的两大类模型———结构化模型与简化形式模型.结构化模型以期权理论解释借款人的还款行为,并以此为主要框架建立贷款定价模型,模型在经济环境波动较大时适应性很好,能准确地捕捉到环境给借款人选择行为带来的变化,但是模型结果不能很好地拟合历史数据,并且纯期权理论分析复杂借款人的行为也是不够的;简化形式模型强调在贷款的历史数据中通过统计方法估计出借款人的还款行为特征和建立定价模型,模型得到的结果比较好地拟合了历史数据,但是,在一个新的经济环境下,模型的样本绩效较差.除这两类主要定价模型外,本文还简要分析了其他几个研究模型的特点.Beginning with studying methods, this paper introduces two types of foreign pricing model-structure model and reduced-form model. Based on option theory that describes borrower＇ s behavior, structure model constructs the pricing model. Under the heavy fluctuant environment, the structure model can rightly capture the change of borrower＇ s behavior, but pricing result does not perfectly fit the historical law, and also it＇ s not sufficient for option theory to explain borrower＇ s behavior. On the contrary, by estimating the characteristics of borrower＇ s behavior from historical data, the reduced-form model constructs the pricing model, and the result can fit the historical law. However, under a new different environment, the model＇ s performance out of sample is very poor. In addition, this paper briefly analyses some other models＇ characteristics.

关 键 词：贷款定价 提前支付 违约 

分 类 号：F830.9[经济管理—金融学]