采用统计学中新近发展的局部线性逼近方法对利率期限结构动态NS模型进行改进,提出了基于局部线性逼近的动态NS模型;并实证比较了改进后的模型与原模型的样本内拟合效果和样本外预测能力。结果表明,改进后的模型无论是样本内拟合效果,还是样本外预测能力都明显优于原模型。
Using the local linear approximation method, this article improves the dynamic NS model and raises the dynamic NS model based on the local linear approximation. Through collecting and analyzing data from China's market, the authors compare the improved model with the original one in the sample fitting effect and prediction ability. The results show that the new model is much better in the sample fitting and prediction effect.