两因素利率模型的参数估计-应用统计专业论文.docxVIP

II II Abstract Interest rates as one of the most basic economic variables, is the basis for asset pricing, financial product design, hedging and risk management, arbitrage and investment. With the rapid development of Chinas bond market and interest rate liberalization, there are more and more researches on the term structure of interest rates. And it is a basic problem to estimate parameters of the interest rate models in the financial sector. This article focuses on the EMM estimation method of two- factor continuous-time stochastic volatility model, in which using the SNP density as the EMM estimated auxiliary model. The paper considers EMM estimation of the two- factor model using weekly observations on the U.S. 3-month T-Bill rate, in which assumed the data obey the two- factor model, and then estimate the model’s parameter. From the estimated results, the two- factor interest rate models don’t well fit the selected data. To further study the effect of the EMM estimates, the paper does EMM estimation of the model using the data from the Monte Carlo method simulation to the two- factor model, the results show that the EMM estimation works well. Finally, to contrastive analysis the effect of parameter estimations of the two-factor interest rate model using the GMM method and EMM method, this paper use the Monte Carlo method to simulate 500 sets of data given a set of model parameters, and then estimate the parameters corresponding to each set of data using EMM method, at last draw the distribution of the parameter estimates. Compare with the result of Travis RA Sapp (2009) which using GMM method, it shows that the effects of the GMM and EMM have their own advantages and disadvantages, but overall EMM method is preferences. Key Words：Interest Rates, Stochastic Volatility Model, EMM, SNP, Monte Carlo Method I III 目 录 摘 要 ··········································································································I Abstract ······················