In his recent paper entitled “Information Uncertainty and Analyst Forecast Behavior”, Frank Zhang explores the effects of an increase in information uncertainty (from either volatility of underlying fundamentals or poor information) on the behavior of sell-side stock analysts. He hypothesizes that if behavioral biases cause analysts to underreact to new information when revising their forecasts, they underreact even more as information uncertainty increases. Using dispersion in analysts’ earnings forecasts as a proxy for information uncertainty over the period 1983-2001, he determines that:
- Upward (downward) forecast revisions predict positive (negative) forecast errors and subsequent additional revisions in the same direction, and forecast errors and subsequent revisions are more pronounced for firms with greater information uncertainty.
- Analysts tend to walk up (down) their estimates for good (bad) news firms as the forecast horizon shortens.
- The effect of information uncertainty on analyst behavior is much stronger following bad news than following good news. Information uncertainty may be greater when management is sitting on bad news, which they release reluctantly.
In summary, stock analysts exhibit predictable underreactions in revising earnings forecasts. The degree of underreaction increases with the earnings forecast dispersion.
