基于遗传算法的多项式样条函数利率期限结构模型  

The Interest Rate Term Structure Model of Polynomial Spline Function Based on Genetic Algorithms

作　　者：白小莹[1] 周荣喜[2] 杨丰梅[1] 

机构地区：[1]北京化工大学理学院,北京100029 [2]北京化工大学经济管理学院,北京100029

出　　处：《系统工程》2009年第7期22-27,共6页Systems Engineering

基　　金：国家自然科学基金资助项目(70701003)

摘　　要：针对样条函数利率期限结构模型中通常根据经验选取分界点的缺陷,本文利用遗传算法寻找多项式样条函数利率期限结构模型的最优分界点,给出了基于遗传算法的多项式样条函数利率期限结构模型,并与固定分界点的多项式样条函数模型进行实证比较。从定价误差看,前者不管是对于样本内的数据还是样本外的数据都优于后者。最后系统地分析了模型与算法中参数:多项式次数、分界点数目、染色体数目和遗传代数对定价误差的影响。Directed at the location of polynomial pieces selected by experience in the polynomial spline interest rate term structure model, this paper employs genetic algorithms to choose the optimal location of polynomial pieces and proposes the polynomial spline interest rate term structure model based on genetic algorithms. Then, an empirical comparison is made between two polynomial spline interest rate term structure models, one is based on the genetic algorithms and the other has fixed location of polynomial pieces. The results show that the errors of pricing of the former are smaller than the latter whether the data are in-sample or out-of-sample. Finally, we are analyzew the influence of the time of polynomial function, number of spline knots, chromosome number and genetic generation on pricing errors in the model.

关 键 词：利率期限结构 遗传算法 多项式样条函数 参数分析 

分 类 号：F830[经济管理—金融学]