我国国债期限利差与经济周期预测研究

发布时间：2018-01-24 18:25

本文关键词： 经济周期经济衰退概率产出缺口期限利差收益率曲线　出处：《复旦大学》2013年硕士论文　论文类型：学位论文

【摘要】：理论和实践都表明,期限利差对经济周期具有很强的预测能力。在经济发展趋势日趋复杂化的今天,对未来适当时期内的经济发展状况进行预测,从货币政策及其他宏观调控政策制定的角度来说意义重大。本文对我国国债期限利差在经济周期预测方面的作用及表现进行了全面研究。首先,对1994至2012年我国实际季度GDP进行HP滤波法分解,得到产出缺口的估计值,并据此对我国经济进行周期测定。测定结果与实际情况和滞后指数所反映情况较为相符。接下来,本文对我国国债市场和期限利差进行详细介绍,并选定10年期国债平均收益率作为长期利率、2年期和1年期国债平均收益率作为短期利率,从而构造我国的国债期限利差。主体部分是,采用Probit模型,用2001年第1季度至2012年第4季度的我国国债期限利差作为自变量,对当期一直到第8个季度之后的我国经济周期状况进行回归,发现由10年期和2年期国债所构造的期限利差具有对经济周期的预测能力,而由10年期和1年期国债所构造的期限利差则没有预测能力。据此给出了我国期限利差和1年后我国经济衰退概率之间的一一对应关系。对先行指数、采购经理人指数、股票指数和债券指数等宏观经济领先指标的类似回归显示,它们均具有对我国经济周期预测的能力,但是先行指数、采购经理人指数的预测能力集中在1年以内,股票指数和期限利差的预测能力在中期的半年到1年半之间,而债券指数的预测能力则在1年半到2年半之间。而半年到1年半的中期预测期限,对于货币政策制定者来说具有十分大的借鉴意义。将期限利差和领先指标一起加入到回归模型中,发现期限利差具有其他领先指标所不具备的对我国经济周期的预测能力,主要体现在预测期限的长度上,即期限利差在中期的预测能力是其他领先指标不具备的。将美国和德国的期限利差加入到回归模型中,结果显示其系数十分显著,然而系数估计值的符号与理论相反,这说明存在伪相关的可能性,因此不能判断外国期限利差对我国经济周期是否具有预测能力。
[Abstract]:Both theory and practice show that the term spread has a strong ability to predict the economic cycle. At a time when the trend of economic development is becoming more and more complicated, this paper forecasts the economic development in the appropriate period in the future. From the point of view of monetary policy and other macro-control policies, this paper makes a comprehensive study on the role and performance of the term spread of national debt in the forecasting of economic cycle. From 1994 to 2012, the GDP of our country was decomposed by HP filtering method, and the estimated output gap was obtained. The results are consistent with the actual situation and the lag index. Next, the paper introduces the bond market and term spread in detail. The average yield of 10-year bonds is chosen as long-term interest rate, and the average yield of 2-year and 1-year Treasury bonds is taken as short-term interest rate, thus constructing the maturity spread of China's treasury bonds. The main part is as follows. Using the Probit model, the term spread of our national debt from in the first quarter of 2001 to in the fourth quarter of 2012 is used as the independent variable. From the current period to the eighth quarter after the state of our economic cycle regression, it is found that the 10-year and 2-year Treasury bonds constructed by the term spreads have the ability to predict the economic cycle. However, the term spread constructed by 10-year and 1-year bonds has no predictive power. Based on this, the one-to-one corresponding relationship between the term spread of China and the probability of economic recession after one year is given. The similar regression of purchasing manager index, stock index and bond index shows that they all have the ability to predict the economic cycle of our country, but the leading index. The forecasting ability of purchasing managers' index is concentrated in less than one year, and the forecasting ability of stock index and term spread is between half and a half years in the medium term. The bond index's predictive power ranges from one and a half years to two and a half years, and half a year to one and a half years. It is of great significance for monetary policy makers. The term spread and leading indicators are added to the regression model. It is found that the term spread has the ability to predict the economic cycle which other leading indicators do not have, mainly reflected in the length of the forecast period. That is, the term spread in the medium term forecasting ability is not available in other leading indicators. The United States and Germany of the term spread into the regression model, the results show that its coefficient is very significant. However, the symbol of the estimated value of the coefficient is contrary to the theory, which indicates that there is a possibility of pseudo-correlation, so it is impossible to judge whether the foreign term interest rate difference has the ability to predict the economic cycle of our country.
【学位授予单位】：复旦大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：F124.8;F832.51;F224

【相似文献】