中国银行间国债利率期限结构实证研究
发布时间：2018-08-16 08:46
【摘要】：利率是指一定时期内利息额与本金的比率,是各种金融资产定价的基础,同时也是国家进行宏观调控的重要指标之一。利率期限结构是指无风险的国债收益率与到期期限之间的关系,反应的是宏观经济中长短期资金的供求关系。同时,利率期限结构与许多宏观经济变量的联系紧密,两者之间的研究一直是金融领域的热点之一。本文首先对利率期限结构的相关理论进行综述,包括利率期限结构的形成理论、利率期限结构拟合模型的发展以及利率期限结构和主要宏观经济变量之间的关系研究。在此基础上本文以中国银行间国债利率期限结构为中心展开,以Nelson-Siegel模型为工具对我国银行间国债的利率期限结构进行了拟合实证分析,得到NS模型的水平、斜率和曲度三因子。在得到模型三因子后,本文对三因子序列进行了进一步的分析。第一步对模型所得的三因子与传统利率期限结构的代理变量进行对比,分析两者之间的异同点;第二步用一阶自回归模型分别对所得的三因子序列进行模型拟合得到未来的预测值,并对预测值的拟合效果进行评估;第三步运用隐马尔科夫模型对利率期限结构的状态转换特征进行了研究;最后运用线性回归模型对宏观经济变量对收益率曲线的影响进行了研究分析。本文的主要结论有:第一,静态拟合模型Nelson-Siegel模型能够拟合我国银行间国债不同形状的利率期限结构;第二,用NS模型估计出的水平因子、斜率因子以及曲度因子与传统利率期限结构的代理变量具有较强的相关性,因此三因子具有明确的经济意义;第三,一阶自回归模型能够较好地预测短期内的利率期限结构,而随着预测时间的延长,模型的预测效果逐渐变差;第四,HMM模型(状态数为2)结果显示我国银行间利率期限结构在短期存在着状态转换,且主要表现为收益率曲线倾斜度的变化,两个状态下的斜率因子均值分别为-1.87和-3.35。最后,NS模型三因子与本文所选宏观经济变量的线性回归模型结果显示:虽然GDP和CPI与利率期限结构的水平因子具有显著的相关性,但是线性回归模型结果显示GDP和CPI对利率期限结构的水平因子的影响非常轻微;GDP、CPI与银行间隔夜拆借利率与斜率因子都具有显著的相关性,根据线性回归模型结果显示银行间隔夜拆借利率对利率期限结构的斜率因子具有非常明显的影响,当银行间隔夜拆借利率上升,即货币政策由宽松变为紧缩时,斜率因子值变大,倾斜度减小,CPI对斜率因子也有明显的影响,CPI的增长会同样会使利率期限结构的倾斜度减小,而GDP对斜率因子只有轻微的影响;曲度因子与GDP、CPI和货币政策之间不存在显著的相关性,因此无法建立相关的线性回归模型。
[Abstract]:Interest rate refers to the ratio of interest to principal in a certain period of time, which is the basis of pricing various financial assets, and is also one of the important indicators of macroeconomic regulation and control. The term structure of interest rate is closely related to many macroeconomic variables, and the research between them has been one of the hot topics in the financial field. Firstly, this paper reviews the related theories of term structure of interest rate, including the formation theory of term structure of interest rate, the development of term structure fitting model, the term structure of interest rate and the main macro-economy. On this basis, this paper takes the term structure of interest rate of Chinese interbank treasury bonds as the center, uses Nelson-Siegel model as a tool to fit the empirical analysis of the term structure of interest rate of Chinese interbank Treasury bonds, and obtains the level, slope and curvature of NS model. This paper makes a further analysis of the three-factor series. The first step is to compare the three factors obtained by the model with the proxy variables of the traditional term structure of interest rates, and analyze the similarities and differences between the two. The second step is to use the first-order autoregressive model to fit the three-factor series to get the future forecast value, and the forecast value. The third step is to use Hidden Markov Model (HMM) to study the state transition characteristics of term structure of interest rate. Finally, we use linear regression model to analyze the impact of macroeconomic variables on the yield curve. Secondly, the level factor, slope factor and curvature factor estimated by NS model have strong correlation with the proxy variables of traditional interest rate term structure, so the three factors have clear economic significance; thirdly, the first-order autoregressive model can better predict the short-term interest rate structure. Fourthly, the HMM model (state number 2) results show that there is a short-term state transition in the term structure of interest rates in China, and the main performance is the change of the slope of the yield curve, the slope factor mean values in the two states are - 1.87 and - 1.87 respectively. Finally, the linear regression model of NS model and macroeconomic variables selected in this paper shows that although GDP and CPI have significant correlation with the level of interest rate term structure, the linear regression model shows that GDP and CPI have very slight impact on the level of interest rate term structure factors; GDP, CPI and bank spacing. There is a significant correlation between the night lending rate and the slope factor. According to the linear regression model, the bank overnight lending rate has a very significant effect on the slope factor of the term structure of interest rate. When the bank overnight lending rate rises, that is, when the monetary policy changes from loose to tight, the slope factor increases and the slope decreases. CPI also has a significant impact on the slope factor, the growth of CPI will also reduce the slope of interest rate term structure, while GDP has only a slight impact on the slope factor; curvature factor has no significant correlation with GDP, CPI and monetary policy, so it is impossible to establish a relevant linear regression model.
【学位授予单位】：浙江财经大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：F224;F822.0
本文编号：2185486
[Abstract]:Interest rate refers to the ratio of interest to principal in a certain period of time, which is the basis of pricing various financial assets, and is also one of the important indicators of macroeconomic regulation and control. The term structure of interest rate is closely related to many macroeconomic variables, and the research between them has been one of the hot topics in the financial field. Firstly, this paper reviews the related theories of term structure of interest rate, including the formation theory of term structure of interest rate, the development of term structure fitting model, the term structure of interest rate and the main macro-economy. On this basis, this paper takes the term structure of interest rate of Chinese interbank treasury bonds as the center, uses Nelson-Siegel model as a tool to fit the empirical analysis of the term structure of interest rate of Chinese interbank Treasury bonds, and obtains the level, slope and curvature of NS model. This paper makes a further analysis of the three-factor series. The first step is to compare the three factors obtained by the model with the proxy variables of the traditional term structure of interest rates, and analyze the similarities and differences between the two. The second step is to use the first-order autoregressive model to fit the three-factor series to get the future forecast value, and the forecast value. The third step is to use Hidden Markov Model (HMM) to study the state transition characteristics of term structure of interest rate. Finally, we use linear regression model to analyze the impact of macroeconomic variables on the yield curve. Secondly, the level factor, slope factor and curvature factor estimated by NS model have strong correlation with the proxy variables of traditional interest rate term structure, so the three factors have clear economic significance; thirdly, the first-order autoregressive model can better predict the short-term interest rate structure. Fourthly, the HMM model (state number 2) results show that there is a short-term state transition in the term structure of interest rates in China, and the main performance is the change of the slope of the yield curve, the slope factor mean values in the two states are - 1.87 and - 1.87 respectively. Finally, the linear regression model of NS model and macroeconomic variables selected in this paper shows that although GDP and CPI have significant correlation with the level of interest rate term structure, the linear regression model shows that GDP and CPI have very slight impact on the level of interest rate term structure factors; GDP, CPI and bank spacing. There is a significant correlation between the night lending rate and the slope factor. According to the linear regression model, the bank overnight lending rate has a very significant effect on the slope factor of the term structure of interest rate. When the bank overnight lending rate rises, that is, when the monetary policy changes from loose to tight, the slope factor increases and the slope decreases. CPI also has a significant impact on the slope factor, the growth of CPI will also reduce the slope of interest rate term structure, while GDP has only a slight impact on the slope factor; curvature factor has no significant correlation with GDP, CPI and monetary policy, so it is impossible to establish a relevant linear regression model.
【学位授予单位】：浙江财经大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：F224;F822.0
【参考文献】
相关期刊论文 前10条
1 陈映洲;张健;;基于动态Svensson模型的国债利率期限结构实证分析[J];统计与信息论坛;2015年04期
2 赵晶;张洋;丁志国;;利率期限结构特征的拟合与预测[J];统计研究;2015年02期
3 丁志国;徐德财;李雯宁;;宏观经济因素影响利率期限结构的稳定性判别[J];数量经济技术经济研究;2014年09期
4 何晓群;王彦飞;;中国利率期限结构与宏观经济运行的关系——基于动态Nelson-Siegel模型的研究[J];经济理论与经济管理;2014年08期
5 陈守东;王妍;;利率期限结构与宏观经济——基于动态潜在因子模型的研究[J];吉林大学社会科学学报;2014年02期
6 孙丽;孙佳佳;;中国同业拆借市场利率期限结构与宏观经济变量关系的实证研究[J];会计与经济研究;2013年05期
7 文忠桥;;中国银行间国债市场利率期限结构实证分析——基于Nelson-Siegel模型[J];财贸研究;2013年03期
8 袁靖;薛伟;;中国利率期限结构与货币政策联合建模的实证研究[J];统计研究;2012年02期
9 潘敏;夏庆;张华华;;货币政策周期与国债利率期限结构[J];财贸研究;2012年01期
10 孙皓;石柱鲜;;中国利率期限结构对经济周期波动的预测能力检验——基于动态Probit模型的研究途径[J];经济与管理研究;2011年04期
，本文编号：2185486
本文链接：https://www.wllwen.com/jingjifazhanlunwen/2185486.html
