Term Structure Of Interbank Bond Interest Rates Based On Vasicek Model Empirical Study

Interest rate reflects the supply and demand level of the supplier and the buyer in the capital market,so interest rate reflects the price level of funds to a certain extent.Term structure of interest rate refers to the price of funds with different maturities.It is a set of interest rates corresponding to funds with different maturities.Because the debt is less risk of default than other financial assets,it can obtain the change information of the benchmark interest rate of different period by the change of the transaction price of the national debt,and get the term structure of the interest rate of the benchmark interest rate market.The term structure of interest rate not only has an impact on the price of treasury bonds,but also provides theoretical support for the price determination of other financial products in the financial market.In addition to determining the price of financial products,the term structure of interest rate plays an important role in the management of portfolio and the management of financial risk in the financial market.In recent years,scholars at home and abroad pay more and more attention to the research on term structure of interest rate,and the research achievements are also increasing.But China's interest rate term structure research started late,and the research results are few.Therefore,this paper attempts to conduct in-depth research on the term structure of interest rate in China's financial market on the basis of previous research results.The first selected data is the daily treasury bond transaction data of the Shanghai stock exchange,and the Nelson-Siegel-Svensson(NSS)interest rate model is used to fit the term structure of the interest rate of the national debt market of the Shanghai stock exchange in China,and the estimated value of the parameters to be estimated for the NSS model is obtained.Then through the construction of AR(2)model,ARMA(1,1)model and RW model to predict the development direction of the model to be estimated,and compare the prediction results of the three models,and then draw the following conclusion: when the duration of the prediction period is short,the prediction effect of ARMA(1,1)model is better than the fitting effect of the ARMA(1,1)model.Therefore,AR(2)model is more suitable for predicting longer interest rates.ARMA(1)model is more suitable for predicting interest rates with shorter maturities.Finally,three economic variables,which are closely related to the term structure of interest rate,are introduced to improve the prediction effect of the model.The estimated parameters of NSS model and three macroeconomic variables are put together,and the rate of interest rate is predicted by stepwise regression method and compared with the traditional forecasting method.The conclusion is as follows: in the long-term prediction,the predictive effect of the stepwise regression method introducing macroeconomic variables is better than the traditional time series prediction method.In the short-term prediction,the traditional time series prediction method is superior to the stepwise regression method introducing macroeconomic variables.