Financial Risk Modelling with Long-Run Weather & Price Variations

Australian farmers face between three and ten times the level of production risk faced by farmers in competing countries worldwide (OECD-FAO, 2011). Consequently, in Australia, identical management plans may result in sharply contrasting distributions of financial outcomes. These differences provide more complete management information than those based on single-point estimates of average outcomes in the average year, which cannot account for the accumulation of financial impacts. This study initially uses data of average monthly rainfalls, average crop, pasture and sheep production, and average input costs and output prices for a representative farm in the Coolamon area of southern NSW (Lat -34.817, Long 147.198, Alt 247m). It compares the profitability of several farming system options involving mixtures of annual dryland crops in rotation with annual and perennial pastures. These data are presented in the documentation of a linear programming (LP) model by Bathgate et al. (2010), we call the ‘Coolamon LP'. A parallel simulation based on the same average-year Coolamon data and performance assumptions, but with long-run series of rainfall data and price variations, a Sequential Multivariate Analysis (SMA) (Hutchings, 2013) was specified with the aim of comparisons with results of the Coolamon LP. The SMA approach simulates multiple 10-year cash balances for the subject farm, after interest and taxes, describing the financial pressures acting upon the business. Results for the SMA model demonstrate the overwhelming importance of downside risk in determining the financial outcomes of farming systems in such dryland farming regions. This demonstrates why an average-season LP model can be a poor, and often incorrect, basis for extension advice on crop and pasture systems or stocking rates