UNIVERSITY OF CALIFORNIA Los Angeles Markov Chain Monte Carlo Estimation of Multi-Factor Affine Term-Structure Models A dissertation submitted in partial satisfaction of the Requirements for the degree Doctor of Philosophy In Statistics by

OF THE DISSERTATION Markov Chain Monte Carlo Estimation Of Multi-Factor Affine Term-Structure Models by He Hu Doctoral of Philosophy in Statistics University of California, Los Angeles, 2005 Professor Jun Liu, Co-Chair Professor Yingnian Wu, Co-Chair This dissertation develops a Bayesian state-space model of the term structure of interest rates. We propose a hybrid Markov Chain Monte Carlo (MCMC) method in estimating multi-factor Vasicek and CIR models with zero-coupon bond yields and interest rate swap data. Inference on the posterior distributions of both model parameters and state variables can be made from the Monte Carlo samples. In contrary to conventional econometric methods, this approach allows the maximum flexibility in allowing measurement errors for all maturities, while it still provides an exact filtering of state variables. The algorithm generates unbiased and efficient estimates in simulation examples. Empirical results show that our approach produces smaller fitting errors with U.S. interest rate swap data than references using different methods.