How certain firm-specific characteristics affect the accuracy and dispersion of analysts' forecasts. A latent variables approach

We suggest that analysts' uncertainty in predicting earnings is a function of: (1) the uncertainty due to production, investment, and financing (PIF) activities, and (2) the amount of information available about the firm. We use a latent variable approach to explore this framework. Observed indicator variables are used to represent the underlying unobserved attributes. The results show that analysts' uncertainty is a positive function of business risk, financial risk, and ownership concentration, and is negatively related to the amount of information.