中文摘要：

分别基于零息票债券的连续复利到期收益率（简称零息收益率）、零息票债券的半年复利平价收益率（简称平价收益率）和瞬时远期利率,对一类轨道随时间非减的Lévy过程――从属过程进行了参数估计和模型之间的比较.对于转移函数已知的从属过程,用极大似然估计得到了参数的估计.对于转移函数未知的从属过程,可以利用鞍点逼近的方法得到转移函数的近似表达式,从而利用极大似然函数进行参数估计.Gamma过程对三类收益率曲线的描述上优于稳定从属过程和泊松随机积分过程.相比较零息收益率和平价收益率,稳定从属过程和泊松随机积分过程在对瞬时远期利率的描述上要吻合一点.

英文摘要：

A number of typical subordinators, which are nondecreasing L6vy processes, are estimated and compared for the treasury yield curve in terms of zero-coupon yield, par yield, and instantaneous forward interest rates. The maximum likelihood estimations for the models with known transition densities are obtained. For models with unknown transition densities, the estimators are derived by saddlepoint approximation method. The Gamma process is successful in capturing the moving of the yield curve. The instantaneous forward interest rate fits the a-stable subordinator and the Poisson random integral better than the zero-coupon yield and the par yield.