Stock Returns and the Miller-Modigliani Valuation Formula : Revisiting the Fama-French Analysis

AbstractFama and French (2006) use the dividend discount model to develop the joint role of three variables – expected profitability, expected investment and current BM – in predicting future stock returns. One reported empirical result is anomalous. The valuation model establishes that the comparative static relation between returns and expected investment is negative, yet it appears to be positive and insignificant. We show that the posited valuation relations between expected returns and the three variables apply at the firm level, and not at the per-share level at which they were tested. Once the variables are measured at the firm level, all the Fama-French predictions are validated