Working Papers
Local Polynomial Estimation of Time-Varying Parameters in GMM (Job Market Paper)
Abstract: In this paper we propose a new nonparametric GMM estimator in the presence of time-varying parameters. We estimate the true time-varying parameters by polynomial approximation fitting as in Fan and Gijbels (1996) and Kristensen and Lee (2019). We show that our proposed estimator retains the properties of consistency and asymptotic normality of the standard GMM under the assumption of uniform locally stationarity. In Monte Carlo study, the proposed estimator shows good performance under various cases of interest such Moving Average model and ARCH model. In the application, we study a simple gravity model for international trade with varying parameters for the US economy and main trade partners and find evidence of varying effects of importer GDP and distance on US exports.
Time-varying Functional Local Projection
Optimal Forecasting with Weighted Least Squares Estimation of Autoregressive Models
Modelling and Forecasting money demand: divide and conquer with Cesar Carrera (Central Reserve Bank of Peru)
Publications
US Monetary Policy Shocks Transmission to Latin America: A GVAR approach. 2016. Revista Estudios Económicos, Banco Central de Reserva del Perú, issue 32, pages 35-54. (In spanish)