Abstract
In explaining the poor informational value of analysts' long-term earnings growth forecasts, studies have focused on the excessively aggressive forecasts induced by analysts' incentives and/or cognitive biases. This study reveals that the forecasts' poor informational value is driven by analysts' reluctance to issue conservative forecasts, which may also be induced by their incentives and/or biases. We predict that this reluctance allows each firm's conservative forecast to be heavily influenced by the firm's past performance and to be the noisy predictors to distinguish high-growth firms from low-growth ones. Consistent with our prediction, we find that each firm's most conservative forecasts are those most strongly influenced by past performance and have the least predictive power.
