Understanding the role of supply and demand factors in the global wheat market: a Structural Vector Autoregressive approach

We present a Bayesian structural Vector Autoregressive model of the global wheat market to examine the relative importance of supply and demand shocks, which are interpreted as the fundamental driving forces of wheat price. To our knowledge, this is the first SVAR analysis that jointly considers (i) a Bayesian non-recursive specification, (ii) production and inventories as endogenous variables (iii) and an inventory-based detection strategy. Our main results indicate that: (i) the posterior median estimates for the price elasticity of supply and demand are mostly similar in their order of magnitude but opposite in signs (0.19 for supply and -0.20 for demand); (ii) the price and the inventories respond to global wheat market shocks differently, depending on the type of structural shock. We also show that the results obtained from Cholesy-type identified annual SVAR models for wheat market are potentially misleading and difficult to reconcile with the economic theory of competitive storage. Finally, we illustrate how unpredictable shifts in supply and demand contributed to the dynamic of wheat price between 2000 and 2022.