DOI: https://doi.org/10.52903/wp2022296
FORECASTING MACROECONOMIC INDICATORS FOR EUROZONE AND GREECE: HOW USEFUL ARE THE OIL PRICE ASSUMPTIONS?
George Filis
University of Patras
Stavros Degiannakis
Bank of Greece
Zacharias Bragoudakis
Bank of Greece
Abstract
This study evaluates oil price forecasts based on their economic significance for macroeconomic predictions. More specifically, we first use the current state-of-the-art frameworks to forecast monthly oil prices and subsequently we use these forecasts, as oil price assumptions, to predict eurozone and Greek inflation rates and industrial production indices. The macroeconomic predictions are generated by means of regression-based models. We show that when we assess oil price forecasts, based on statistical loss functions, the MIDAS models, as well as the futures-based forecasts outperform those generated by the VAR and BVAR models. By contrast, in terms of their economic significance we show that none of the oil price forecasts are capable of providing predictive gains for the eurozone core inflation rate and the Greek industrial production index, whereas some gains are evident for the eurozone industrial production index and the Greek core inflation rate. However, in all cases the oil price forecasting models, including the random-walk, generate equal macroeconomic predictive accuracy. Thus, overall, we show that it is important to assess oil price forecasting frameworks based on the purpose that they are designed to serve, rather than based on their ability to predict oil prices per se.
Keywords: Oil price forecasts, MIDAS, conditional forecasts, core inflation, industrial production
JEL-classifications: C53, E27, E37, Q47
Acknowledgement: We thank the anonymous reviewer and Hiona Balfoussia for their constructive suggestions which helped us to improve the clarity of the paper. The views expressed in this paper are those of the author and do not necessarily reflect those of the Bank of Greece. This research was conducted in the context of the Bank’s programme of cooperation with universities.
Correspondence:
George Filis
University of Patras
Department of Economics
University Campus, 26504, Rio