Title: Time-varying SURE: A nonparametric approach
Authors: Isabel Casas - University of Southern Denmark (Denmark) [presenting]
Eva Ferreira - University of the Basque Country (Spain)
Susan Orbe - University of the Basque Country (Spain)
Abstract: A local linear estimator is proposed for a SUR model (LL-SUR) with time-varying coefficients. The asymptotic results, consistency and asymptotic normality, are obtained under locally stationary variables. Contrary to the parametric case, the results show that the efficiency of LL-SUR outperforms the LL estimator separately estimated for each equation even when the covariates are the same for all equations. A simulation study is conducted to show the performance of the methodology in finite samples. Moreover, a discussion and practical procedure is derived to select the smoothing parameters. Finally, we present an application to estimate time-varying coefficients in beta pricing models.