No abstract

No abstract

No abstract

Stochastic computable general equilibrium (CGE) models have ignored regional correlations in agricultural yields, assuming random shocks to be independent between regions. This could lead to misinterpretation of simulation outputs which ignore extreme positive or negative harvests at the global scale. We develop a multi-regional CGE model which allows for five types of interregional correlation between wheat yields to analyse the vulnerability of countries against fluctuating international markets, focusing on Value at Risk (VaR) and extreme dependency. We find that global welfare risks could be underestimated by up to 33% if significant interregional correlations in yield shocks are not taken into account. Egypt, Kazakhstan, Ukraine, the former Soviet Union and Northern Africa are particularly vulnerable to global volatilities in terms of economic welfare.

No abstract