债券隐含利率与多因子Vasicek模型

Modeling the Term-Structure of Interest Rates in the SSE with Multiple-Factor Vasicek Model

以上海证券交易所(简称上交所)债券价格隐含的利率期限结构,从1996-06～2003-02的周样本数据作为分析对象,首先利用主成分分析法对利率期限结构的变化进行分析,发现需要2～3个状态变量,利率模型才可能反映利率期限结构的变化。然后实证研究了连续时间的两因子、三因子Vasicek模型对上交所利率期限结构的描述情况。结果表明,两因子Vasicek模型可以反映样本期内上交所利率期限结构的形状,但模型不能反映利率期限结构的时间序列变化。同时发现,三因子模型相对于两因子模型,不能够明显改进对上交所利率期限结构的拟合。

With the weekly data of term structures of interest rates from june 1996 to february 2003 in the Shanghai stock exchange, principal component analysis approach is used to study how many state variables of interest rate model are suitable to model the term structure changes in the SSE, and it indicates 2 or 3 factors are needed. With Kalman filter and maximum likelihood estimation approaches, two-factor, three-factor continuous-time Vasicek model is estimated. Empirical study indicates that two-factor model models the relative changes of the term structures very well, but two-factor model is not so good in modeling the time series changes of the term-structure. It is also found that compared with two-factor, three-factor Vasicek model don't improve the fit significantly.