The article compares traditional and copula-based methods for measuring risk in agricultural portfolios. The researchers found that using copulas to model non-Normal returns can create more accurate risk measures like Value at Risk (VaR). By applying Gaussian and Student's t copula models to different asset classes, they were able to create portfolios that offer better protection against losses.