Proceedings of The 7th International Conference on Advanced Research in Business, Management and Economics
Inflation Forecast in G7 Economies: The Role of The High-Frequency Data of Energy Prices
Helena Chuliá Jorge A. Muñoz-Mendoza1, Jorge M. Uribe
We study the predictive power of the oil and other fossil fuel prices as high-frequency predictors for monthly inflation in the G7 countries. The data covers the period between January 2, 2011 and June 30, 2022. For this purpose, we use LASSO-MIDAS models and other conventional models for inflation forecasting exercises. Our results show that the LASSO-MIDAS models have better predictive performance than the other models such as MIDAS, VAR and ARIMA. These results have better accuracy and significance in countries such as Canada, France, Germany, and the United Kingdom, both in in-sample and out-of-sample exercises. Energy prices have greater predictive power compared to the metals and food prices, in addition to having time-varying behavior. These results have important implications for policymakers, regulators, and forecasters due to their implications for economic and financial policy decisions.
keywords: energies, forecasting, LASSO-MIDAS, forecast combination