摘要
本文针对LGD模型开发中的核心问题进行了理论研究和实证分析,根据中国银行业的实际,建立在清收基础上的历史平均LGD是一种较为实用的方法,根据这一方法论,本文运用决策树建立了LGD模型,模型不仅考虑了回收成本的因素,而且考虑了时间因素。本文的实证研究结果表明,不同抵押的债项的LGD与新资本协议确定的参数基本接近,总体上来看,时间因素对于LGD的影响要较回收成本的高。若不考虑回收成本,LGD的估值要低约1%;而忽略时间价值,LGD估值要低约7%。分析LGD的影响因素发现:除债务类型、行业及信用评级等因素对LGD有较大影响外,区域因素的影响不可忽视,建立LGD模型时有必要加以考虑。
Based on requirements of Basel II and domestic researches, this paper conducts a theoretical and empirical study on several core issues in LGD development. Due to the soundness of applying methodology of workout historical average LGD in China, we build a LGD model that takes into consideration not only recovery cost but also discount rate. The result shows that LGD with different collaterals are approximate to the parameters determined by Basel II, but discount rate has a higher effect on LGD rather than recovery cost, without considering recovery cost, LGD may be underestimated about 1%, but 7% if not considering discount rate. And it also finds that except for debt type, industrial and credit rating, location has a relatively higher effect on LGD, thus must corporate this factor into the LGD models in China.
出处
《国际金融研究》
CSSCI
北大核心
2009年第6期71-78,共8页
Studies of International Finance
