多因素时变Markov链模型下考虑信用风险的互换期权定价  

The valuation of swaption with counterparty risk under a multi-factor time-varying Markov chain model

作　　者：陈正声[1,2] 秦学志[2] 

机构地区：[1]大连银行风险管理部,大连116001 [2]大连理工大学管理与经济学部,大连116024

出　　处：《系统工程理论与实践》2011年第6期993-1003,共11页Systems Engineering-Theory & Practice

基　　金：国家自然科学基金(70771018);教育部人文社会科学基金(05JA630005)

摘　　要：提出具有似扩散行为的共同因素与特定评级因素及其驱动下的时变Markov链模型,在仿射期限结构框架下建立了考虑交易对手信用风险的互换期权定价模型.以1998年1月-2008年12月美国国债的收益率数据及评级机构穆迪的信用评级数据为样本,利用卡尔曼滤波技术与约束非线性最小二乘法对模型进行了参数估计.主要结果为:首先,以似扩散过程的形式引入具有权重影响的共同因素与特定评级因素,构建了基于信用评级的时变Markov链模型;其次,建立了含信用风险的互换期权定价模型,给出了该期权的闭式解;最后,分析了交易对手方信用等级状况的变化对互换期权定价的影响.This paper suggests a time-varying Markov chain model with weighted common and rating specific factors as diffusion-like behavior in affine term structure framwork,which is applicable to pricing of swaption with counterparty risk.Based on monthly yield data of US Treasury for the period of Jan 1998 to Dec 2008 and Moody＇s rating data,Kalman filter and constrained nonlinear least square are used to estimate the models.The main results of this paper are：First,the model with common and rating specific factors as diffusion-like process is introduced.Second,the valuation model of swaption with counterparty risk is studied and the closed form solution is also obtained.Finally,we show that the impact of credit standing of counterparty on the price of swaption.

关 键 词：互换期权 交易对手风险 信用评级 时变Markov链模型 

分 类 号：F830[经济管理—金融学]