Abstract: Contrary to conventional wisdom, growth stocks (low book-to-market stocks) do not have substantially higher future cash-flow growth rates or substantially longer cash-flow durations than value stocks, in both rebalanced and buy-and-hold portfolios. The efficiency growth, survivorship and look-back biases, and rebalancing effect help explain the results. This finding suggests that duration alone is unlikely to explain the value premium. Using rebalanced portfolios, I find that, consistent with asset pricing models that feature countercyclical risk premiums, there is a growth premium in the cross section of stock returns. That is, risky assets with higher expected cash-flow growth rates have higher expected returns, after controlling for cash-flow risks.