Price transmission in European fish markets
Résumé
We investigate price transmission in European fisheries markets. To start, we identify clusters of fish species both cross-regionally and within countries. A major issue in the clustering exercise is missing data due to reporting issues and seasonality in landed fish catches. To handle this we implement k-POD clustering which unlike traditional k-means clustering is able to account for missing data in clustering. Next, we move on to our primary goal of investigating price transmission through modelling price volatility spillovers. A missing value VAR framework is applied to identify directional spillovers among fish prices within clusters. We show that the directional volatility spillover effect is more prominent within each country for different species clusters than that for species cross-regionally. This suggests price transmission within localized markets for fish, rather than cross-regional markets for individual species. Although there are some neighbour country effects and some species that appear to be priced cross-regionally. Our study is the first to explore these multiple fish pricing dynamics, particularly taking account of the important issue of missing data which is a notable feature of fish pricing data.