Abstract

This paper first describes the main features of supply and demand in cocoa spot markets. A state- variable model is proposed to describe the random evolution of cocoa forward curves over time, which essentially adapts to agricultural commodities, introduced by Borovkova and Geman (2006) for energy. In contrast to most of the literature on the subject, the first state variable is not the spot price, as it combines seasonal and stochastic features and may not be observable, instead, the average value of all liquid futures contracts is a quantity devoid of seasonality and conveys a robust representation of the forward curve level. The second state variable is a quantity analogous to the stochastic convenience yield, which accounts for the random changes in the shape of the forward curve. We conduct estimation procedures for the cocoa market over the period of 1980 to 2009 and exhibit an interesting result on cocoa seasonality as well as an extension of the Samuelson effect.

Key words: Cocoa Spot markets, forward curve stochastic modeling, future curve average value.