Time-varying nonlinear regression models: Nonparametric estimation and model selection

TL;DR

This paper develops nonparametric methods for estimating and selecting models of time-varying nonlinear regression functions in time series, with theoretical guarantees and practical application to interest rate data.

Contribution

It introduces an information criterion for model selection in nonparametric time-varying nonlinear regression and proves its consistency.

Findings

Asymptotic theory for estimates of time-varying regression functions

Proposed information criterion is selection consistent

Application to U.S. Treasury interest rate data

Abstract

This paper considers a general class of nonparametric time series regression models where the regression function can be time-dependent. We establish an asymptotic theory for estimates of the time-varying regression functions. For this general class of models, an important issue in practice is to address the necessity of modeling the regression function as nonlinear and time-varying. To tackle this, we propose an information criterion and prove its selection consistency property. The results are applied to the U.S. Treasury interest rate data.