100,0o0 yields per product (i.e., 100k potential scenarios). Probability distributions that best fit the his- torical data of product yield factors were selected", and a multivariate random sample of 100,000 factor for each product was created using these distributions and the correlation matrix presented above. These sim- ulated risk factors were used to simulate Yields per product, Volume of production based on an expected area, and the Value of the production based on a reference price*. This approach enabled the assessment of the Joint Probability Distribution of the Portfolio Value. 55 The Weibull distribution best fts wheat, maize, rape and sunflower; while the Pareto distribution is most appropriate for barley, green maize and soya. In turn, a Pareto with left heavy tail was selected for milk. 56 The number of productive units (hectares or animals) and the price per ton are treated as fixed multiplicative constants.