Coffee prices in Brazil: predictive variables in cash and future markets

Abstract

Literature describes categories of variables directly and indirectly affecting cash and future coffee prices in addition to the historical prices and other variables related to offer and demand. Behavior of cash (IPMF) and future (IPFU) coffee prices were analyzed in relation to stocks (EST) and exports (EXP) to identify predictive information useful for the Brazilian market. Monthly data from January 1996 to July 2005 were used to estimate an Auto-regressive Vector model (VAR) with posterior variance decomposition and accumulated responses to impulses of the variables IPMF, IPFU, EST and EXPV. In general the conclusions permit identification of predictive information indicating: a) the relative importance of the variance of coffee exports in explaining cash price variance; b) the importance of cash price variability in explaining future prices, and not the contrary; c) that a shock in endogenous variables analyzed caused a similar behavior both in the IPMF and IPFU; and d) that a shock in future coffee prices caused stocks to increase slightly, indicating an incentive for speculation in this commodity.