Abstract: A statistical model for calculating farm profit and economic risk was presented in view of the influence of various factors on farm crop yield and agricultural income. On the basis of the existing statistical theory, the model used the conditional value-at-risk （CVaR） to evaluate the effectiveness of geographical diversification. CVaR was used to benchmark losses, while the Copula function was used to model the joint distribution of marginal revenues. The results showed that geographic diversification could be a feasible agricultural risk management method for wheat farm portfolio managers, which could achieve the optimal expected returns while controlling the risk. Compared with the traditional multivariate normal distribution model, the average CVaR model based on Copula could better simulate extreme losses, providing an innovative solution for agricultural risk management.

Key words: farm profit, geographical diversification, conditional value-at-risk （CVaR）, Copula functions, portfolio