The Empirical Studies On Nonparametric Estimation Methods Of Interest Rate Term Structure Models Based On Chinese Short-term Interest Rate

Interest rate is an important economic indicator. With the development and globalization offinancial markets, the research on interest term structure models is of great significance. Thedevelopment of external financial market is earlier and more perfect, while the domesticfinancial market is still relatively lack of the study on the short-term interest rate term structuremodels, although it develops rapidly since reform and opening up. Most scholars use parameterestimation theoryto study short-term interest rate term structure models in the existing literature,and they limit the flexibility of drift and the diffusion in rate models by setting the function form,therefore it is likely to neglect some dynamic characteristics of rate. At present, few scholars setthe function form by using nonparametric estimation.It is necessary to analyze and estimateChinese short-term interest rate flexibly as well as find reasonable interest rate term structurediffusion model.This paper choose7day bonds repo rate in the Shanghai exchange treasury as the sampledata, obtain five traditional parameter models by GMM estimation, and get the nonparametricestimation model by Ait-Sahalia method. The parameter estimation models and nonparametricestimation modelare analyzed by comparing drift, diffusion and marginal densitiesto assess theempirical application ability. At last, we get a nonparametric interest rate term structure modelwhich grasps the dynamic characteristics of interest rate accurately.