Probabilistic Modeling of the Coffee Market in Brazil

Abstract

The coffee is one of the most valuable commodities in the world and in this scenario, Brazil presents itself as the largest producer and exporter in the world. However, the high fluctuations in prices promote insecurity in the agents of the sector. In this sense, the objective of this work is to propose the best probabilistic model for the monthly analysis of prices and to calculate probabilities of occurrences of average prices of coffee bags according to levels of practical interest. To this end, we use historical data provided by COOXUPÉ corresponding to the period from January 1981 to December 2022, arranged into monthly subseries. For supporting the results, goodness of fit test were performed. The results indicated that the Gamma and log-Normal distributions fit the coffee bag price data in all months. The log-Normal distribution outperformed in all months. The Gamma and log-Normal distributions fitted the monthly data of average coffee bags prices. The Log-Normal distribution is more suitable for the probabilistic study of the variable in all months. January, February, and March are the months with the highest probability of higher average values and are therefore the most recommended for coffee trading.