Speculative price bubbles are defined as a significant deviation between an asset's intrinsic value and its market value and in this paper it refers to stock values. Literature about the theme has noted the existence of bubbles in various types of markets and their respective assets. A great deal of effort has been directed toward identifying bubbles in stock price indices. However, few research endeavors focus on assets as the unit of analysis. Studies about stocks in Brazil have identified the presence of bubbles in IBOVESPA (São Paulo Stock Exchange Index). Given this context and assuming that the speculative bubbles are present in the Brazilian stock market, this research is focused on the following question: Is there evidence of the existence of speculative bubbles in stock prices traded on the São Paulo Stock Exchange? Econometric tests were performed on twenty-seven stocks, based upon their positions each semester, for the period between the first semesters of 1990 until the first semester of 2010. The nominal values of the selected stocks were adjusted for inflation by the IPCA (Brazilian Consumer Price Index). In order to identify the presence of bubbles, we applied the Johansen non-cointegration test and/or the Granger non-causality test between the intrinsic value, dividends and interest on equity capital, and the market value (semester closing price) of the stocks. The primary findings reveal a presence of bubbles in twenty of the twenty-seven stocks, at a

Este artigo tem o objetivo de testar a forma fraca de eficiência do mercado futuro brasileiro da commoditie agrícola café arábica, usando a técnica de cointegração a fim de verificar se os preços futuros correntes são estimadores não viesados dos preços à vista esperados para o futuro. Para isso, utiliza as séries históricas de janeiro de 2005 a maio de 2011 dos preços futuros, que foram coletados na Bolsa de Mercadorias e Futuros -BM&F, e os preços à vista calculados pelo CEPEA/ESALQ/USP. As métricas utilizadas são os testes ADF de Dickey e Fuller para detectar a presença de raiz unitária e o teste de cointegração de Johansen para verificar a existência de um relacionamento de longo prazo. Os resultados indicaram a não estacionariedade das séries de preços e a presença de cointegração. Porém, o teste dos parâmetros α = 0 e β = 1 da regressão que comprovam a eficiência fraca e não viés encontraram indícios estatísticos de não eficiência de mercado, bem como da presença de um viés indicando a existência de um prêmio associado ao risco.

No abstract