Comparing Autoregressive vs. Markov Switching- Autoregressive Models in Fish Meal Price Forecasting

The objective of this paper is to present a parsimonious forecasting model for the fishmeal price. The focus is on the soybean meal market’s impact on the fish meal price through the soybean meal futures price together with the stocks-to-use as an indicator of demand and supply conditions. A salient feature of the fishmeal market is the impact of the ENSO events on fishmeal supply. This possibly leads to two different price regimes, one where the fish meal price is highly correlated with the soybean meal price, and another, during ENSO events, where fish meal supply is low and the fish meal price is not strongly correlated with the soybean meal price. The results from the Markov-switching autoregressions indicate two price regimes where one is mostly governed by the soybean meal price while the other is governed by the level of stocks-to-use.