Agricultural commodities markets provide an important venue for the transfer and pricing of large volumes of key inputs of food production. The structure of the market and the activity and motivations of market participants are factors in pricing stability and ultimately the costs for food produced and consumed on world markets. This paper uses a modified behavioural finance model and trader categories

defined by the Commodity Futures Trading Commission to examine sentiment driven price changes and speculative activity in wheat, soybean, rough rice, soybean meal, soybean oil, corn and live cattle. The paper further combines a Vector Autoregression (VAR) model and Cumby-Modest tests to describe the positive feedback trading behaviour of different investors. The analysis shows that speculators are positive feedback traders and that, while previous returns influence sentiment, sentiment does not influence returns except when sentiment is high