Olive Oil World Price Forecasting: A Bayesian VAR Approach

Abstract The Tunisian olive oil strategy is based on the development of exports. Extension and modernization measures of Tunisian olivegrove have come into effect these recent years in order to increase the exports and diversify the target markets. Like any other agricultural good, olive oil is subject to world price fluctuations. Forecasting the long-term world price of olive oil is essential both as a decision-making tool and as a strategic factor for the development of the sector. This paper attempts to forecast the long-term olive oil world price using annual time series. The data reveals that the number of observations is too restrictive for a frequentist approach. To overcome the sample shortage, we adopt a Bayesian VAR. We use the hierarchical prior selection to specify the prior parameters. The results show an increase in world price and production. However, the price grows more proportionally than the production. In such a context, the increasing production orientation seems adequate but not sufficient to enhance the Tunisian position on the international olive oil market. For the coming decade, the Tunisian olive oil policy should focus as well on deepening the dynamics of product valorization.