短期利率波动测度与预测:基于混频宏观-短期利率模型

Measuring and Forecasting Short Rate's Volatility:Based on Mixed Frequency Short Rate Model with Macro Factor

经验研究表明宏观经济对识别短期利率及其波动现象具有重要影响。为合理引入宏观信息并精确拟合与预测短期利率波动行为,本文提出一种包含宏观因子的混频短期利率模型,即BHK-MIDAS模型。基于中国宏观及利率数据信息的研究结果表明:与传统短期利率模型相比,BHK-MIDAS模型具有更优的样本内拟合效果;相对于货币政策指标而言,宏观基本面与通胀指标对短期利率波动的贡献更大;进一步地,混频模型还可以识别出受宏观因子显著影响的短期利率波动的长期成分;特别地,BHK-MIDAS模型在短期利率波动样本外预测方面的良好表现,充分说明宏观因子在识别及预期短期利率波动行为方面的重要贡献。

Many empirical studies show that macroeconomic play important rule in short rate and its volatility behavior. For better fitting and forecasting the volatility of short rate with macroeconomic information, this paper proposes the mixed frequency short rate model with macro factors, namely BHK - MIDAS model. Based on the Chinese data we find that： compared with the traditional short rate model, BHK - MIDAS model exhibits the better in - sample performance. The macro fundamental and price indicator contribute more for the volatility of short rate than that of monetary policy indicators. Furthermore, mixed frequency model can identify the time vary long term component affected by macro factors. Particularly, BHK - MIDAS model presents the better out -sample performance too. It is implied that the macro factors contribute substantially to identification and prediction of short rate＇ s volatility.