This paper examines the hypothesis of asst return predictability in the
Brazilian Stock Market (Bovespa). Evidence suggests that seven factors
explain most of the monthly differential returns of the stocks included in
the sample. Within the factors that present statistically significant mean,
two are liquidity factors (market capitalization and trading volume trend),
three refer to price level of stocks (dividend to price, dividend to price
trend, and cash flow to price), and two relate to price history of stocks (3
and 12 months excess return). Contradicting theoretical assumptions, risk
factors present no explanatory power on cross-sectional returns. Using an
expected return factor model, it is contended that stock returns are quite
predictable. An investment simulation shows that the model is able to
assemble portfolios with statistically significant higher returns.
Additional tests indicate that the winner portfolios are not fundamentally
riskier suggesting mispricing of assets in the Brazilian stock
Market.