Macroeconomic uncertainty and quantitative vs. qualitative inputs to analyst risk forecasts

We study how sell-side analysts map both quantitative and qualitative information from earnings conference calls into their forecasts of fundamental firm risk. We find that analysts perceive firm risk to be lower when absolute earnings surprises are small and tone of the earnings conference calls is more positive. Further, we find that the relative importance of qualitative information increases during periods of high macro-uncertainty (i.e., during NBER crises or periods of high VIX), and this increase improves the calibration of the risk forecasts. While our main results rely on a general measure of Tone of language to gauge the qualitative information component of the earnings conference call, additional analyses find that both positive and negative language affects risk forecasts and that extreme language rather than moderate language results in differential risk perceptions. Our results are robust to alternative empirical specifications and increase our understanding of the “black box” that is the analyst forecasting process.