【作者】 林强；

【导师】 冯玲；

【作者基本信息】 福州大学 ， 金融工程， 2020， 硕士

【摘要】 利率期限结构又称收益率曲线,国债利率期限结构与经济增长存在紧密的关系。国债收益率曲线的特征会反映在斜率上,假定期限差不变,可以用长短期国债远期期限利差来简单表示国债远期收益率曲线的斜率。因此,若能找到适用于我国实际经济情况的远期利率期限结构模型,进而模拟国债远期利率期限结构和长短期国债远期期限利差,则可以据此来预测经济增长的走势,对充实我国国债利率期限结构理论研究亦有参考价值。为了提高利率期限结构对经济增长的预测精度,可以从利率期限结构模型创新和用于预测的数据源两方面寻求突破。在利率期限结构模型创新方面,本文引入量子场论下HJM模型,用纳入不同到期日远期利率相关性的单个二维量子场代替一维白噪声来重构远期利率的演化方程,以解决多因子HJM模型待估参数繁多的问题;在经济增长预测的数据源方面,本文将利用2011-2017年国债远期利率期限结构估计模型参数,并模拟出2018年的日频国债远期利率期限结构,避免了宏观—金融模型使用低频数据会损失信息的问题,检验样本期内国债远期利率期限结构对经济增长的预测作用,最后根据模拟的国债远期利率期限结构来预测经济增长的走势。首先,本文利用国债远期的即期收益率日频数据,构建应用广泛的传统两因子HJM模型、量子场论下HJM模型。通过均方根误差(RMSE)、Theil不等系数对比两个模型的预测精度发现,不管是国债远期利率期限结构还是长短期国债远期期限利差,量子场论下HJM模型的预测精度都要优于传统HJM模型。其次,实证分析刻画国债收益率曲线斜率的长短期国债远期期限利差与GDP、工业增加值累积增长率的关联性。实证结果表明:长短期国债远期期限利差与GDP、工业增加值累积增长率呈现显著的正相关关系。最后,检验量子场论HJM模型模拟的长短期国债远期期限利差对经济增长预测作用的稳健性,发现量子场论模型更稳健。综上,本文构建的量子场论HJM模型模拟的国债远期利率期限结构对经济增长的预测能力更佳,为经济增长预测提供了一种新的模型,有助于根据国债收益率曲线特征来预测经济增长的走势。更多还原

【Abstract】 The term structure of interest rate is also known as yield curve.The term structure of interest rate of Treasury bonds is closely related to economic growth.The characteristics of the Treasury yield curve will be reflected in the slope.Assuming that the term difference is unchanged,the slope of the Treasury forward yield curve can be simply expressed by the spread of long-term and short-term Treasury bonds.Therefore,if we can find a term structure model of forward interest rate that is applicable to the actual economic situation of our country,and then simulate the term structure of forward interest rate of Treasury bonds and the spread of long-term and short-term Treasury bonds,then we can predict the trend of economic growth accordingly,and it is also of reference value to enrich the theoretical research on the term structure of interest rate of our country.In order to improve the forecasting accuracy of the term structure of interest rate on the economic growth,we can seek a breakthrough from the innovation of the term structure model of interest rate and the data source used for forecasting.In terms of the innovation of term structure model of interest rate,this paper introduces the HJM model under the quantum field theory,and uses a single two-dimensional quantum field with the correlation of different maturity forward rates to replace the one-dimensional white noise to reconstruct the evolution equation of forward rates,so as to solve the problem of many parameters to be estimated in the multi-factor HJM model.In terms of economic growth forecast data source,this article will use the national debt forward term structure of interest rates of 2011-2017 to estimate model parameters,and simulate the daily frequency data of national debt forward term structure of interest rates of 2018,to avoid the problem that macro-financial models lose information by using low-frequency data,test sample period bond forward term structure of interest rates forecast effect on economic growth,according to the simulation of bond forward term structure of interest rates can be used to predict the trend of the economic growth.Firstly,the paper constructs the widely used traditional two-factor HJM model and the quantum field theory HJM model by using the daily frequency data of the spot yield of Treasury bonds.By comparing the prediction accuracy of the two models with the root mean square error(RMSE)and Theil inequality coefficient,it is found that the prediction accuracy of the HJM model under the quantum field theory is better than that of the traditional HJM model,regardless of the term structure of the forward interest rate of Treasury bonds or the spread of the long-term and short-term Treasury bonds.Secondly,empirical analysis is made on the correlation between long-term and short-term Treasury term spreads and the cumulative growth rates of GDP and industrial added value,which characterizes the slope of Treasury yield curve.The empirical results show that there is a significant positive correlation between the spread of long-term and short-term Treasury bonds and the cumulative growth rate of GDP and industrial added value.Finally,by testing the robustness of the effect of the long and short term Treasury term spread simulated by the HJM model of quantum field theory on the prediction of economic growth,it is found that the quantum field theory model is more robust.To sum up,the term structure of Treasury forward rates simulated by the quantum field theory HJM model in this paper has a better prediction ability for economic growth,which provides a new model for economic growth prediction and helps to predict the trend of economic growth according to the characteristics of Treasury yield curve.更多还原