Here, the control variables are similar to the ones in the accuracy model (i.e., Equation
(1)), but we do not include number of analysts (NANA), loss dummy (LOSS), and earnings
level (EL) because the conceptual link between these variables and forecast dispersion is
not clearly established in the literature.14 We expect that large (SIZE) firms would have
smaller dispersions, while financially distressed (ZMIJ) firms, firms with longer forecast
horizon (HORIZON), and firms with more volatile earnings streams (SURPRISE and
STDROE) would have larger dispersions.