摘要
波动结构对于可违约债券及其衍生品的定价和风险管理具有重要意义。利用AAA级企业债券价格数据,基于中国可违约债券市场构建三因子可违约随机波动HJM模型,并对其进行有限维马尔科夫仿射实现。在此基础上,从波动因子的随机波动特征、相关性结构和贡献度3个方面对中国可违约债券市场的波动结构进行系统分析。研究结果表明,样本期内中国可违约债券隐含的无风险利率和信用利差的波动率中含有显著的随机波动过程,且其数值呈现持续增大的趋势;无风险短期利率、短期信用利差和随机波动过程3个主要波动因子之间存在显著的相关关系;各波动因子的风险贡献度随时间推移而发生明显的波动。在经济向好时期,无风险短期利率的风险贡献较大;在经济趋冷时期,短期信用利差的风险贡献占优。
Volatility structure is of significance in pricing and hedging of defanltab｜e bonds and its derivatives. Using AAA corpo- rate bond price data, the term structure of defauhable bonds in China is first of all calculated by the extended Nelson-Siegel mod- el based on genetic algorithm, and then analyzed by econometric methods, such as principal componem analysis, Jarque-Bera test and ARCH LM test. It shows that the term structure of defaultable bonds is mainly driven by three volatility factors with a significant ARCH characteristic from the perspective of dynamics. In the light of above analysis above, a defauhable HJM model with stochastic volatility is established by decomposing the defauhable short rate into the default-free short rate and the short-term credit spread, and setting their volatilities dependent at both state variable＇s level and in the stochastic volatility process. A finite dimensional affine realisation is obtained for the established model under appropriate volatility specifications. Then, a comparison is carried out between the established model and the HJM model with level-dependent volatility, demonstrating that the estab- lished model offers a much better performance in both data fitting and distribution depicting. Based on the established model, a systematic analysis is made on the volatility structure of the defauhable bond market in China in the three aspects of the stochastic volatility feature, correlation structure and contribution of the volatility factors. The result shows that the volatilities of both default-free interest rate and credit spread implied in the defauhable bond price have sig- nificant stochastic volatility, indicating a trend of continuous increase in the sample period with a considerable influence on the bond pricing precision and error distribution. There exist significant correlations among the three main volatility factors of the de- fault-free short rate, the short-term credit spread and the stochastic volatility. It is noteworthy that the defauh-free sho
出处
《管理科学》
CSSCI
北大核心
2015年第1期122-132,共11页
Journal of Management Science
基金
国家自然科学基金(71171144)
教育部人文社会科学研究青年基金(11YJCZH147)
高等学校博士学科点专项科研基金(20130032120013)
教育部长江学者和创新团队发展计划项目(IRT1028)~~
