Crespo Cuaresma, Jesus Jumah, Adusei Karbuz, Sohbet
Year of Publication:
Working Papers in Economics and Statistics 2007-22
We propose a new time series model aimed at forecasting crude oil prices. The proposed specification is an unobserved components model with an asymmetric cyclical component. The asymmetric cycle is defined as a sine-cosine wave where the frequency of the cycle depends on past oil price observations. We show that oil price forecasts improve significantly when this asymmetry is explicitly modelled.
oil price forecasting nonlinear time series analysis asymmetric cycles