This analysis helps identifying the risk and uncertainty linked to some of the variables used in agricultural market projections© EU, 2013
Taking account of uncertainty in agricultural market projections - JRC analysis
The OECD–FAO 2013 Agricultural Outlook for the period 2013-2022 has been released yesterday. It projects a slow-down in the medium-term expansion of world agricultural production because of limited area expansion and slower productivity growth. Agricultural commodity prices in real terms will be held above pre-2007 levels by tight market conditions and higher input costs. Agricultural trade is expected to increase, with most of the export growth coming from developing countries.
The JRC has contributed to this influential annual report for the second time, drawing on its recent partial stochastic analysis of agricultural market projections. This kind of analysis attempts to discern the relative importance of key specific risks. The approach does not attempt to forecast the implications of all possible uncertainties for future market outcomes. Instead, it allows the policymaker to select specific sources of uncertainty and to quantify the likely range of variation around the projected results that derives from them. This approach involves performing multiple simulations (up to 500) in order to quantify the implications for the market outlook of uncertainty about future values of key market drivers like exchange rates, energy prices and crop yields. This work provides a complex picture of how the risk associated with these factors is transmitted to production volumes, commodity prices and trade flows, thereby adding a crucial dimension to the policy-relevant information provided by the outlook.
Partial stochastic analysis has already been used by the JRC in preparing the annual market outlook of the Directorate-General for Agriculture and Rural Development, which also uses the AGLINK-COSIMO model. The technique is being further developed within the JRC’s agri-economic modelling platform, iMAP, in order to extend it to other iMAP models and increase its user-responsive flexibility.
Ex ante policy analysis explores the medium-term impacts of future policy changes, which are measured against a ‘no-policy-change’ reference scenario. The European Commission publishes an annual medium-term outlook for the main agricultural commodity sectors, based on the partial equilibrium model AGLINK-COSIMO, which projects EU supply balance sheets (production, consumption, exports, and imports) 8-10 years ahead.
These projections serve as the reference scenario for ex ante simulations of EU agricultural policy changes. Although the baseline is not a forecast, it is nevertheless often interpreted by policy makers and market analysts as an indication of the most likely future market trends given the implemented, or already adopted, policy remains unchanged. Partial stochastic analysis can enhance the value of baseline information by setting it in a probabilistic context that recognises some of the risk and uncertainty underlying the projected values. This information supplements the point estimate provided by the deterministic (non-stochastic) baseline, and allows the user to take into account the relative uncertainty of the different projected variables.
The JRC has just published a new reference report dealing with the methodology of JRC's partial stochastic analysis used when preparing the annual market outlook of the Directorate-General for Agriculture and Rural Development or the recently released OECD–FAO Agricultural Outlook for 2013-2022.