This paper aims to test the hypothesis of abnormal returns from the strategy of investing in shares with lower Price/Value Ratio (PVPA) in the Brazilian capital market. All the shares negotiated in the Sao Paulo Stock Exchange (Bovespa) from 1994 to 2006 were used, and formed 6 portfolios according to the original criteria for choosing portfolios, in ascending order of PVPA, changed every year. Moreover, we tested the existence of significant change in the parameters of the CAPM Model through the regression analysis, incorporating to the model the variable corresponding to multiple PVPA. Afterwards, we also made a comparison between the governments of FHC and the first Lula's administration, checking if there was significant change in the parameters of the regression, using the Chow Test of structural change. The results showed that there isn't efficacy in the use of low PVPA as a measure of investment portfolio formation, rejecting the hypothesis tested both through the graphic analysis, which showed higher returns for the shares with greater PVPA, and through the inclusion of this index in CAPM, which indicated significant change on the beta parameters and in the comparison between the governments.
PVPA index; Bovespa; Market anomalies; Data panel regression; Investment myth
