Search

We examine the association between dividends and price through the lens of the Ohlson (Contemp Account Res 11:661–687, 1995, Contemp Account Res 18:107–120, 2001) accounting valuation framework. Employing the price-level model and coefficient definitions detailed in Ohlson (Contemp Account Res 18:107–120, 2001), we show that in a properly specified pricing model a positive dividend coefficient does not violate Miller and Modigliani (J Bus 34:411–433, 1961), imply market irrationality, nor signify an informational role for dividends. Using a simple illustration, we explain the intuition behind the positive dividend coefficient. We also demonstrate that in the price-level model Ohlson's "other information" variable νt, is better specified as the expected change in earnings—not just the earnings forecast. Our results hold through a number of robustness tests. [ABSTRACT FROM AUTHOR]