Shibor Interest Rate Jump-diffusion Model Error Correction Threshold Estimation

The term structure of interest rates is a hot topic in modern finance research.The jump-diffusion model is one of the effective models for describing short-term interest rates.In this paper,the kind of error correction threshold estimation methods is used to estimate the volatility coefficient of the model:the asymmetric kernel estimation reduces the asymptotic variance of the estimation to reduce the corresponding error.In the simulation experiment In the simulation experiment,it can be found that the error correction threshold estimation can reduce the estimation error more effectively than the traditional estimation method,so as to provide a theoretical basis for accurately measuring the diffusion coefficient of the interest rate.Empirical analysis,this paper mainly Shibor interest rate from 2007 to 2017 as the object of study,we can find the error correction threshold can effectively detect Shibor interest rate jump behavior,the other with the simulation experiment on the Shibor interest rate volatility for non-error The threshold is revised to further identify and measure interest rate risk and provide ideas and methods for better understanding the characteristics of China's interest rate market.By comparing the research status at home and abroad,the contributions of this paper mainly lie in the following three aspects: First,the asymmetric estimation error correction threshold estimation method is used,and the comparison between the symmetrical kernel estimation method and the asymmetric kernel estimation method is finally performed.Asymmetric kernel estimation reduces the asymptotic variance of the estimator and reduces the coincidence error.Second,the choice of threshold factor and the choice of window width.Building the GARCH(1,1)time-varying threshold function avoids the use of the same threshold value for financial data with different periods of continuous performance(high volatility at a certain time and low volatility at a certain time)Estimate or underestimate jumping behavior.The theoretical value method and CV(Cross Validation)method are used to select the optimal window width in nonparametric estimation.Third,the previous scholars mainly studied foreign financial markets.What we are focusing on here is China's financial market.The simulation results and empirical data were used to explore and compare the effects of the wave method.The estimation method is universal and can be used not only for stationary time series but also for non-stationary time series.