| Forecasts of long-term earnings growth (LTG) are important for valuation. Prior research employs firm-level models to forecast LTG. However, research in economics, management, and accounting is unclear as to whether industry-level models would be more useful. I study whether firm- or industry-level models produce more accurate LTG forecasts. I investigate when industry-level forecasts are more accurate than firm-level forecasts, which forecast serves as the best proxy for market expectations and whether the difference between the industry and firm forecast is predictive of future returns. Industry-level forecasts of LTG are more accurate than firm-level forecasts and are more accurate than the median I/B/E/S consensus analyst long-term growth forecast, a zero growth forecast, a forecast combining industry and firm information and forecasts from the Li (2003) and Hou et al. (2012) models. The industry-level forecast is most accurate than the firm-level forecast for extreme firm-year observations, for firms with highly variable earnings and for firms with a large market share. The industry-level forecast is less accurate for firms in the Decline life cycle stage. Finally, I find that the difference in the industry and firm-level forecasts is predictive of future returns, suggesting that market prices do not fully incorporate the implications of industry-level financial statement information for long-term earnings growth. |
