An International Meeting of the American Accounting Association American Accounting Association 2006 Annual Meeting August 6–9, 2006 Washington, D.C.

Jing Li
Columbia University

Abstract: This study examines how accounting conservatism affects analysts’ earnings forecast, taking into account the interaction between unconditional conservatism and conditional conservatism. Since more unconditional conservatism leads to less uncertainty about earnings response to new information, analysts’ forecasts will be more accurate and less biased. Empirically, I find that absolute forecast error is smaller for firms with more unconditional conservatism. Unconditional conservatism has greater impact on analysts’ forecast accuracy for early forecast when uncertainty is greater than for late forecast. The improvement in forecast accuracy of bad news firms due to unconditional conservatism is larger than good news firms. For the analysts’ forecast bias, I find partial evidence that the bias in both pessimistic and optimistic forecasts decreases as firms become more unconditional conservative.

Back to Session Listing

AAA Home Page