Incorporating uncertainties within the FAPRI-UK Modelling System: A stochastic approach
Siyi Feng1, Julian Binfield2, Myles Patton1 and John Davis1
1) Agri-Food & Biosciences Institute and 2) University of Missouri
The FAPRI-UK partial equilibrium modelling system has been developed to capture the key uncertainties associated with agricultural production systems using a stochastic framework1. The deterministic FAPRI-UK model generates single point estimates for prices, livestock numbers etc. These point projections are based on normal weather conditions, specific macro-economic and other exogenous assumptions. In contrast, the stochastic modelling system incorporates uncertainty in the future paths of key variables, generating 500 sets of projections based on alternative world agricultural commodity prices, oil prices, exchange rates, crop yields and demand. The process of generating stochastic projections involves estimating distributions among certain exogenous variables and error terms in the equations and simulating the models 500 times with respect to the resulting stochastic draws. By varying assumptions about certain exogenous variables, stochastic models can be used to examine the different ways markets may behave.