Nonparametric estimation of multifactor continuous time interest rate models

This paper studies the finite sample properties of the kernel regression method of Boudoukh et al. (1998) for estimating multifactor continuous-time term structure models. Monte Carlo simulations are employed, with a grid-search technique to find the optimal kernel bandwidth. The estimator exhibits truncation and correlated residuals biases near the boundaries of the data. However, the variance of the estimator is so high that the biases are unlikely to be relevant from a hypothesis testing point of view. The performance of the estimator is also studied under model misspecification. Irrelevant regressors reduce efficiency and induce additional biases in the estimates. Using Treasury bill data, I test whether the estimates produced by the nonparametric estimator are statistically distinguishable from estimates obtained under a parametric model. The kernel regressions pick up nonlinearities in the data that the parametric model cannot capture.