摘要
文章从国债利率期限结构的无套利动态Nelson-Siegel模型估计方法入手,对卡尔曼滤波算法进行改进,以提升模型参数估计的精确度;同时提取模型中的动态因子,考察其与宏观经济信息之间的关联性;进一步利用我国国债2010—2021年债券交易数据进行实证研究。结果表明:基于改进估计算法的模型无论是样本内预测还是样本外预测,效果均优于传统卡尔曼滤波估计的结果,同时发现状态因子与宏观经济变量之间有较强的关联性,蕴含了较为丰富的宏观信息。
This paper starts with the estimation method of no-arbitrage dynamic Nelson-Siegel model of interest rate term structure of treasury bonds to improve the Kalman filter algorithm so as to increase the accuracy of model parameter estimation.At the same time, the dynamic factors in the model are extracted to examine the correlation between the dynamic factors and macroeconomic information. Finally, the empirical study is carried out based on the bond transaction data of China’s treasury bonds from 2010 to 2021. The results show that both in-sample prediction and out-of-sample prediction, the effect of the model based on the improved estimation algorithm is better than that of the traditional Kalman filter estimation. Simultaneously, it is found that there is a strong association between state factors and macroeconomic variables, containing rich macroeconomic information,which has a guiding significance to our macroeconomic policies.
作者
李桂芝
张奇松
王雪标
Li Guizhi;Zhang Qisong;Wang Xuebiao(School of Economics,Dongbei University of Finance and Economics,Dalian Liaoning 116000,China;School of Economics and Management,Yingkou Institute of Technology,Yingkou Liaoning 115000,China;School of Information and Business Management,Dalian Neusoft University of Information,Dalian Liaoning 116023,China)
出处
《统计与决策》
CSSCI
北大核心
2022年第19期124-130,共7页
Statistics & Decision
基金
国家自然科学基金面上项目(71273044)
辽宁省博士启动基金一般项目(2020-BS-269)
营口理工学院创新团队支持计划(TD201903)。
