Copper return to predictability in the frequency domain

Abstract

This dissertation aims to assess the enhancement of out-of-sample predictability in copper returns through the application of frequency domain techniques. For this purpose, a wavelet method is employed - Maximal Overlap Discrete Wavelet Transform Multiresolution Analysis (MODWT MRA), based on the research of (Faria and Verona, 2020). Additionally, our predictor variables are grounded in the work of (Gargano and Timmermann, 2014; Zhang et al., 2021), expanding the literature on out-of-sample forecasts, wavelet-based methods (MODWT MRA), and out-of-sample copper predictability. The results obtained demonstrate both statistical and economic gains when implementing frequency domain techniques to predict out-of-sample copper returns. These findings align with recent studies on out-of-sample forecasting of equity returns (Faria and Verona, 2018, 2020, 2021).

Date of Award	16 Jul 2024
Original language	English
Awarding Institution	Universidade Católica Portuguesa
Supervisor	Fábio Verona (Supervisor) & Gonçalo Faria (Co-Supervisor)