Is Analyst Optimism Intentional? Additional Evidence on the Existence of Reporting and Selection Bias in Analyst Earnings Forecasts
54 Pages Posted: 26 Mar 2001
Date Written: February 2001
The paper examines whether the documented bias in analyst earnings forecasts is intentional by investigating whether analysts account for the market's ability to adjust for the bias when forming their forecasts. The procedure follows three steps. First, I estimate on a firm-specific basis the forecast error for each firm decomposed in its two parts, bias and underreaction. I then employ an ERC model for each firm modified to include the expected bias and underreaction from step 1. The coefficient on the expected bias variable serves as the proxy for the market's ability to adjust for the bias. Finally I include this proxy in a cross-sectional regression explaining analyst forecast error. Finding a relationship between bias and the proxy for the market's ability to adjust for the bias will provide evidence on the joint hypothesis that (a) analysts know what the ability of the market to undo the bias for a firm is and (b) they form or issue their forecasts accordingly. A positive relationship is consistent with the existence of both reporting and selection bias. A negative relationship is consistent with the existence of reporting bias.
I find that analysts reduce the bias in their announced forecasts as the market's ability to adjust for the bias increases. This result provides direct evidence that analysts knowingly bias their forecasts. The negative relationship between the market's ability to adjust for the bias and the level of bias in the forecasts provides support for the existence of reporting bias in particular. Controlling for the sign of bias (optimistic vs. pessimistic) shows that optimistic bias is intentional. Even though I do not find evidence that pessimistic bias is intentional this may be due to the structural design of the tests. Finally, even though I find that the market seems to adjust for analyst underreaction, I do not find evidence that analyst underreaction is intentional.
Keywords: Analyst forecasts; Bias; Optimism; Pessimism; Underreaction; Reporting bias; Selection bias
JEL Classification: M41, G29, G14
Suggested Citation: Suggested Citation