Template-Type: ReDIF-Paper 1.0
Series: Tinbergen Institute Discussion Papers
Creation-Date: 2024-12-31
Number: 24-082/III
Author-Name: Pierluigi Vallarino
Author-Workplace-Name: Erasmus University Rotterdam and Tinbergen Institute
Title: Dynamic kernel models
Abstract: This paper introduces the family of Dynamic Kernel models. These models study the predictive density function of a time series through a weighted average of kernel densities possessing a dynamic bandwidth. A general specification is presented and several particular models are studied in details. We propose an M-estimator for model parameters and derive its asymptotic properties under a misspecified setting.
A consistent density estimator also introduced. Monte Carlo results show that the new models effectively track the time-varying distribution of several data generating processes. Dynamic Kernel models outperform extant kernel-based approaches in tracking the predictive distribution of GDP growth.
Classification-JEL: C14, C51, C53
Keywords: 
File-URL: https://papers.tinbergen.nl/24082.pdf
File-Format: application/pdf
File-Size: 1.066.438 bytes
Handle: RePEc:tin:wpaper:20240082