Geographical diversification in wheat farming: a copula-based CVaR framework

Purpose - Portfolio theory suggests that geographical diversification of production units could potentially help manage the risks associated with farming, yet little research has been done to evaluate the effectiveness of a geographical diversification strategy in agriculture. The paper aims to discuss this issue. Design/methodology/approach - The paper utilizes several tools from modern finance theory, including Conditional Value-at-Risk (CVaR) and copulas, to construct a model for the evaluation of a diversification strategy. The proposed model - the copula-based mean-CVaR model - is then applied to the producer's acreage allocation problem to examine the potential benefits of risk reduction from a geographical diversification strategy in US wheat farming. Along with the copula-based model, the multivariate-normal mean-CVaR model is also estimated as a benchmark. Findings - The mean-CVaR optimization results suggest that geographical diversification is a viable risk management strategy from a farm's profit margin perspective. In addition, the copula-based model appears more appropriate than the traditional multivariate-normal model for conservative agricultural producers who are concerned with the extreme losses of farm profitability in that the later model tends to underestimate the minimum level of risk faced by the producers for a given level of profitability. Originality/value - The effectiveness of geographical diversification in US wheat farming is evaluated. As a methodological contribution, the copula approach is used to model the joint distribution of profit margins and CVaR is employed as a measure of downside risk.