We reexamine signaling and agency theories and argue that the free-cash-flo
w hypothesis implies a stronger information effect for both over- and under
investing firms than for value-maximizing firms. Our results indicate that
dividend and capital structure policies interact to provide significant pre
dictive information about future cash flow. We also find a U-shaped relatio
n between the amount of information and Tobin's q. The minimum of this rela
tion occurs near a q value of one. This outcome implies a stronger informat
ion effect for both over- and underinvesting firms than for value-maximizin
g firms.