Abstract
It is a well-felt recent phenomenal fact that global food prices have dramatically increased and attracted attention from practitioners and researchers. In line with this attraction, this study uncovers the impact of global factors on predicting food prices in an empirical comparison by using machine learning algorithms and time series econometric models. Covering eight global explanatory variables and monthly data from January 1991 to May 2021, the results show that machine learning algorithms reveal a better performance than time series econometric models while Multi-layer Perceptron is defined as the best machine learning algorithm among alternatives. Furthermore, the one-month lagged global food prices are found to be the most significant factor on the global food prices followed by raw material prices, fertilizer prices, and oil prices, respectively. Thus, the results highlight the effects of fluctuations in the global variables on global food prices. Additionally, policy implications are discussed.