Weak-Form Market Efficiency of the Brazilian Exchange Rate: Evidence from an Artificial Neural Network Model

Abstract

This article utilizes an artificial neural network model to examine the hypothesis of weak-form market efficiency for the monthly Brazilian exchange rate from 1999 to 2013. The method of partial derivatives suggested by Racine and White (2001 Racine, J., & White, H. (2001). Statistical inference, the bootstrap, and neural network modeling with application to foreign exchange rates. IEEE Transactions on Neural Networks, 12(4), 1–19.[PubMed] , [Google Scholar]) is used. The first step is to choose network architecture, second, weights estimation, and, at last, testing according to the suggested procedure. The results suggest that the Brazilian foreign exchange market is not efficient informally; thus, agents can obtain unusual profits through arbitrage.