中文摘要：

以上海证券交易所的国债回购利率数据为样本，本文采用两种不同核函数：高斯核和抛物线核对非参数利率期限结构模型进行估计．结果显示：短期利率的密度函数是非正态的，扩散过程的漂移函数和扩散函数都是非线性的，高斯核比抛物线核对扩散函数拟合更平滑．然后，给出了基于非参数和参数利率模型的国债定价的方法，并对非参数利率模型、Vasicek模型、CIR模型、多项式样条静态模型进行国债定价预测比较与分析．

英文摘要：

With the data of the repurchasing rate in Shanghai stock market, the nonparametric term structure model of interest rates is estimated by using two different kernel functions： Gauss kernel function and Epanechnikov kernel function in this paper. The empirical results show that the density function of short interest rate is non-normal distribution, and the drift and diffusion function are nonlinear, and Gauss kernel function is smoother than Epanechnikov kernel function for fitting diffusion function. Then the method of pricing treasury based on nonparametric and parametric interest rate models are given, and the forecasting of treasury prices of nonparametric interest rate model, Vasicek model, CIR model and polynomial splines static model are compared and analyzed.