An Analysis on Agricultural Market Behavior

Abstract

This dissertation is concerned with (i) how to model an agricultural market, (ii) how to analyze the impacts of a certain event (i.e. animal disease outbreak) on the market, and (iii) what are the relationships between different markets. The research on the first two issues will focus on the US beef market, and the impact of the bovine spongiform encephalopathy (BSE) outbreak (Dec. 2003) on the US beef market will be analyzed. For the third issue, a multinational meat market will be considered, which includes three countries (Korea, US, and UK) and three meat products (beef, pork, and poultry). Their market movements will be compared, considering the impacts of the major animal disease outbreaks: BSE, foot and mouth disease (FMD), and avian influenza (AI). Based on the properties of an agricultural product (longer cycle of production and perishability) and the extensive empirical results, it is concluded that a recursive model is appropriate for modeling an agricultural market. A variety of structural change tests are applied to reveal that the change due to the BSE event still lies in an allowable range of the prediction error. For the comparisons between market movements, some multivariate statistical methods such as canonical correlation analysis (CCA) and principal component analysis (PCA) are used, and the main finding is that the knowledge about the threat of BSE to human health played an important role in changing people's attitude towards an animal disease event.