Forecasting monthly fisheries prices: model comparison using data from Cornwall (UK)

Forecasting has a key role in applied economics and management of fisheries. In this paper we report the forecasting competition between Autoregressive AR(p), Moving Average MA(q) and ARMA(p,q) models of the monthly average fisheries prices. We consider twelve species landed into Cornwall: Anglerfish, Cod, Crabs, Dogfish, Haddock, Hake, Lemonsole, Mackerel, Plaice, Saithe, Sole and Whiting. In our evaluation of the outof- sample forecasting accuracy of ten models, we show that simple ARMA(p,q) models generally prove to be the best forecasting models.