History Matters : Individual Analysts' Responses to Quarterly Guidance

We ask whether the history of interactions between a firm and its analysts affects management's quarterly earnings forecast decision and analysts' subsequent earnings estimate revisions. We show that the history of their interactions not only reflects their individual incentives but is also used to form expectations about current decisions. Specifically, guidance is more likely to be pessimistically biased if management has a history of offering pessimistic forecasts and if analysts have tended to respond passively (i.e., by mimicking guidance or revising in the direction indicated by management). Individual analysts are more likely to respond passively to management's current forecast if they have historically responded in such a manner or if management has tended to offer unbiased or accurate guidance. Analysts with superior earnings forecasting track records are less likely to respond passively to guidance that is pessimistically biased. Collectively, our results indicate that downward bias in quarterly guidance and information production by analysts depend on what management and analysts have learned about each other over time as reflected in their forecasting histories