Simulating the effect of El Ni\~no Southern Oscillation on the worldwide wheat prices

This work analyses the impact of the large-scale atmospheric-oceanic phenomenon known as El Niño Southern Oscillation on wheat production, and the consequent changes in prices at the global scale by using computer simulations. Several intermediate results are obtained on the way to the final goal. The identification of geographic areas relevant to the international wheat market and the integration of heterogeneous datasets are two of them. Building on these two results, the local effects of the El Niño Southern Oscillation phases on the wheat yield are quantified using robust ANOVA regression, and their potential impacts on the aggregate production of each area are estimated. Finally, these estimates are provided as inputs to the computational model, which outputs, among others, the wheat prices of 12 internationally relevant production areas. Simulation results show how the cross-section distributions of prices, conditional on the occurring of El Niño and La Niña, spread to the right compared to that observed for the neutral phase. Therefore, both non-neutral phases imply an increase of average and dispersion of prices, although the effect of La Niña is weaker than that of El Niño.