Abstract: The correlation and leading relationship between the yield curve of China’s national bonds and the macroeconomic indicators have received widespread attention from macro-control and financial markets, of which the nonlinear and time-frequency characteristics are waiting for extended researches. Based on the yield curve data of the national bonds from 2006 to 2022, this paper employs a dynamic NS model to study the fitting term structure of national bond interest rate. The findings show that the term structure of national bond interest rate presents certain cyclical fluctuation characteristics. With the extension of maturity, the yield curve shows a gradual convergence trend. The quantile vector auto regression model is utilized to study the nonlinear impact of the term structure of the national bond interest rate on the macroeconomic indicators under different economic levels, it is found that the level factor and slope factor of the national bond yield rate mainly have negative effects on output and inflation. When the macro-economy is at different levels, this negative effect has nonlinear characteristics, which is especially greater in the periods of high economic growth and high inflation. The wavelet phase spectrum method is used to explore the dynamic changes of the forecasting ability of the term structure of national bond interest rate to the macroeconomic indicators in the time-frequency dimension. It is found that the horizontal factor and the slope factor have stronger forecasting ability to output, while the forecasting ability to inflation has weakened after 2019. Therefore, in the future, the construction of the national bond market should be promoted, the monitoring of the term structure of the national bond yield rates should be strengthened, and the fiscal and monetary policies should be optimized.

Key words: national bond yield, term structure of interest rate, macroeconomic effect, nonlinear characteristics