时间相依利率扩散模型的非参数估计

The Nonparametric Estimation of Time-Dependent Diffusion Models

利率的变动模式会随着时间的推移、经济环境的变化和金融制度的改变而发生变化。时变的扩散模型能更好的描述短期利率的随机行为,文章采用基于核回归的非参数方法,估计中国银行间市场7天回购利率的时间相依CKLS模型。最后比较了时间相依CKLS模型与时齐CKLS模型在波动率预报上的表现,结果表明,时间相依CKLS模型更好的反映了利率实时的变化,提高了预测利率变化的精度。

The time- dependent diffusion models can better capture the stochastic behavior of short - term interest rates. Nonparametric methods based on kernel regression are used to estimate the time - dependent CKLS model for China interbank market 7 - day repo rates. The forecasting power of the time- dependent CKLS model and that of the time- homogeneous CKLS model are also compared. The results show that time- dependent CKLS model improves the forecasting precision of the interest rates.