国债回购市场利率期限结构的预测能力研究

Research of Predictive Power of Interest Rate Term Structure of Chinese Government Bond Repo Market

本文对国债回购市场的利率数据在采用单位根和协整检验的基础上,通过分析国债回购利率走势和波动幅度的不同,将全部样本数据分成两个子样本,并在基于利率预期假说成立的基础上,利用单方程回归和向量自回归模型来检验收益率价差对未来利率变化的预测能力。结果发现,利率价差在两个子样本区间上都对未来的利率变化有较强的预测能力,且利率波动幅度越大,则利率价差对未来利率变化的预测能力越强;反之,则越弱。

The paper, on the basis of testing government bond repo market＇s interest rate data with unit root and cointegration method, divides all sample data into two sub-samples by analyzing government bond repo interest rate trend and volatility range. Then, based on the existence of EMH, the paper utilizes single-equation regression and vector auto-regression model to test the predictive power of yield spread for the future interest rate change. The result shows that yield spread possesses good predictive power in the two sub samples and the power is higher in the more volatile first sub-sample. Thus, the yield spread in Chinese government bond repo market includes some useful information, reflects the expectation of market for the future interest rate change and may be used as a leading indicator to monitor the effect of conducting monetary policy by the Central Bank.