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

Flora H Zhou
Cornell University

Abstract: Dye and Sridhar (2004) provide a game-theoretic framework for examining the extent to which manipulable information should be incorporated into accounting reports to maximize their value-relevance. Using a variant of their model, I show that investor uncertainty about management incentives to manipulate information reduces the relevance-maximizing weighting on manipulable information in equilibrium. I also predict that greater weight on manipulable information and greater incentive uncertainty are likely to reduce the predictive power of the equilibrium because they increase the game’s “strategic dependence” (Bloomfield 1995). A laboratory experiment confirms these predictions, and the data suggest that fair-value accounting (which places relatively high weight on manipulable information) may harm investor welfare more than predicted by equilibrium analyses, and may also generate predictable fluctuations in the perceived and actual value-relevance of financial reports.

Back to Session Listing

AAA Home Page