Abnormal accounting growth and analyst forecasts

Abstract

Analysts center on a firm's business growth to forecast earnings. The more uncertainty in growth, the more difficult for analysts to forecast. Therefore, forecasts are more dispersed and less accurate. To evaluate this hypothesis, this paper structures the content of uncertainty in growth and integrates them into one metric: abnormal accounting growth (AAG). AAG captures two dimensions of uncertainty in growth: 1) the uncertainty from the disagreement across various growth rates and 2) the uncertainty from the deviation of the firm-specific mean to the grand mean. Validity tests show that AAG materially and incrementally contributes to traditional risk metrics: the volatility of weekly return and market beta. Empirical results confirm that high AAG distinctly explains high forecast dispersion and low forecast accuracy. Besides traditional determinants AAG's explanatory power is incremental. In addition, when earnings surprise is negative or loss occurs, it is more difficult to forecast. Accordingly, the effect of AAG on forecast dispersion and accuracy is magnified. This work emphasizes that 1) the disagreement across growth rates and 2) high or low growth both are risky. Through AAG, this work complements the understanding of how uncertainty in growth affects forecasting performance.