A nonparametric method to estimate time varying coefficients under seasonality: An application to the Spanish money multiplier.

This work analyzes the estimation of a time varying coefficients regression model in presence of seasonality. We propose the use of a nonparametric regression method restricted to a seasonal constraint. The resulting estimator generalizes a wide class of seasonal estimators, being very flexible, consistent and asymptotically normal distributed under very general conditions. Furtheremore, it is able to combinate the smoothness and seasonal restrictions, its advantage is that it only depends on the smoothness assumption about the variability of the parameters. We illustrate its performance by estimating the Spanish money multiplier