Análise de variação de preços utilizando cadeias de Markov e lógica Fuzzy

Abstract

Concepts of Markov chains and Fuzzy logic are important theories that focus on dealing with uncertainties, it can be applied the forecast of price variation of exchange traded products. This paper aims to present an application to the concepts of Markov chains and Fuzzy logic. The data presented in this paper were collected on the Center for Advanced Studies in Applied Economics, University of S˜ao Paulo, for the entire soybean and chicken commodity pricing basis. It discusses how the both theories are considered important, as they may be associated with the study of time series, as well as some of their most interesting characteristics and definitions for the model presented. It may also support other work that is interested in other types of price fluctuations in various exchange or over the counter assets. Considering the size of the export to the Brazilian market, information such as price change analysis is useful for business management. The described models presented very similar results for soybean price variation, clearly presenting the positive trend of this time series. Regarding the variation of chicken prices, the models varied slightly, presenting other possibilities of analysis.