Multivariate functional-coefficient regression models for nonlinear vector time series data

Abstract

SUMMARY Vector time series data are widely found in practice. In this paper we propose a multivariate functional-coefficient regression model with heteroscedasticity for modeling such data. The local linear smoother is employed to estimate the unknown coefficient matrices. Asymptotic normality 10 of the proposed estimators is established, and bandwidth selection is considered. To deal with cointegration widely observed in financial markets, we propose an error-corrected multivariate functional-coefficient model. Simulations show that the proposed estimation procedures capture nonlinear structures of coefficients well. Analysis of United States interest rates illustrates the proposed methodology. 15