The purpose of this paper is to test market efficiency by using the so-called ''technical analysis'' investment strategies. We apply predictive hability tests developed by White (2000) in order to measure whether these strategies are capable of producing excess returns relative to the buy-and-hold strategy. It considers the possibilty of data-snooping, allowing the detection of models whose seemingly good performance are in fact due only to chance. Our results indicate that technical analysis is not profitable and should not be used by investors, being in line with weak-form efficient market hypothesis predictions.
