Empirical Analytic Applications Of Stochastic Interest Rate Models

With the development of the process of marketization of interest rates, and the continued expansion of fixed-income securities in the investment market, so as to have great practical significance to study the dynamic term structure of interest rates.This paper through the theory of stochastic processes to instruct stochastic interest rate models. For Ho-Lee, BDT two typical no-arbitrage stochastic interest rate model, and obtain market price data by converted fixed coupon bonds into zero-coupon bonds, which were constructed interest rate and price binomial model when volatility parameters fixed and variable parameters volatility in assumption, take single goal seeking and solver in Excel to solver nonlinear equations, and use Merton Carlo simulation and verification distribution of the end term structure of interest rates. When using the overnight Shibor data to estimate Vasicek, CIR two equilibrium stochastic interest rate model parameter under a combination WLS method, Merton Carlo simulation method for a coupling variable technical conditions and the conditional variance formula.For Vasicek model prediction error, we use the ARIMA model for data re-fitting and made of economics interpretation to deviations. Under the premise of the CIR model release fixed volatility parameter assumptions, take GARCH(1,1) model combines maximum likelihood estimation method to estimate the variable volatility parameter. By using different period Shibor date to fit the model to verify that the curve shape of the term structure of interest rates. This paper use Excel, Eviews, Spss software and Python languages to select market data, estimate model parameters, analysis the fitting results. It emphasizes operability and practicality simultaneously, through a dynamic model depicts the trajectory of interest rates.