The Study On Dynamic Term Structure Of Chinese Interest Rate Based On MCMC Method

Interest rate is one of the most important and fundamental parameters in financialmarket. It play a fundamental role in both macro economy and micro-financial market.From the view of macro economy, it is the vehicle of the monetary policy, and has avital influence of expectation of inflation as well the target of many governments overthe world. From the view of micro financial market, it plays a fundamental role in assetpricing, risk management. As for the importance of the interest rate, A lot of literaturehave focused on the dynamics of it.Interest rates of almost all countries have the leverage effect, volatilityclustering andvolatilitysmile. To capture these characteristics, researchers have built a lot of models.Based on these literatures, this thesis models the dynamic of interest rate by threedifferent general models, which are deterministic volatility model, stochastic modeland jump-diffusion model. To extend the models so they can describe morecharacteristics, this thesis tries to add leverage effect, jump effect and no-Gaussianeffect to above model. Through these jobs, the fitted rate of models can be improved.Based on the MCMC method,this thesis simultaneously estimates the parameters ofdeterministic model, stochastic volatility model and jump-diffusion model, gives acomprehensive compare among these model under the DIC rule, and finally finds thatChinese inter-bank7-days repo rate represent an obvious leverage effect and volatilityclustering, the jump effect in multi-factor model is not that satisfying and the bestmodel to fit Chinese inter-bank7-days repo rate is SV-Level model.